diff --git "a/main/part_2/0158757491.json" "b/main/part_2/0158757491.json"
new file mode 100644--- /dev/null
+++ "b/main/part_2/0158757491.json"
@@ -0,0 +1 @@
+{"metadata":{"gardian_id":"e411c024840bd43eb6c85cdecffa879b","source":"gardian_index","url":"https://cgspace.cgiar.org/rest/bitstreams/3cffd45d-1a60-4deb-ae89-9cda44548a6b/retrieve","id":"-162103369"},"keywords":["Chicken","Diversity","Ethiopia","Genomic","Indigenous","Population Chicken","Genetic Diversity","Ethiopia","LEI0258 microsatellite","MHC Indigenous","Chicken","Improved Horro","Signature of selection","SNP 4.5. Conclusions Indigenous","Chicken","Improved Horro","Signature of selection","SNP"],"sieverID":"8865e940-9699-4a15-9f23-a6abdd9d1be1","content":"for your unforgettable deeds when I was hopeless to secure the budget for my Ph.D. study.You really dragged me out from my interwoven academic problems and lifted me up to this level.Dr. Biru Yitaferu, your role for my success is unforgettable as the director general of Amhara Regional Institute. Among others, I can underline your honest understanding of my problems which I thank you a lot. My wife, Senait Mekit, I well know that you shouldered a big responsibility in managing the family in my absence. My beloved kids, Petros, Tsion, Aaron, and Aleph Adebabay, I was away from you and didn't contribute a lot as a father in those study years for the best of the family. May God grow you in wisdom and richly bless you. My dear brother Ayalew Kebede, I won't forget your contribution for my development in life. My appreciation also extends to my sister in law, Zufan Yirga and my wonderful nieces Kalkidan, Liya and Bethlehem Ayalew for their kindness and hospitality during my stay as a member of the family.vi My utmost appreciation also goes to Bioscience for Eastern and Central Africa (BeCA) which gave me the platform and good exposure for molecular laboratory Analysis.Alshawi, and Ibrahim Akinci for their mentoring in my stay at the University of Nottingham for the basic exposure to Bioinformatics. Rosalie, Camille, and Brian Hanotte, the fascinating family of Olivier Hanotte, you made my stay in Nottingham very pleasant because of your warm welcome and adorable hospitality. Many other people and institutions which I didn't mention here have played a great role in my success in one way or another. Many thanks all! I cannot conclude my acknowledgment without thanking people working at different levels in agricultural offices of sampling districts for their kind cooperation for the data collection process.Farmers were also very helpful to me during the sampling process, thanks a lot.                            4. Genome statistics for the chicken reference genome (Gallus Gallus 5.0) assembly. .................... 33 Table 5. An overview of common approaches for detecting signatures of selection. The number of alleles, allele frequency, and heterozygosity levels were used to measure diversity within populations whilst the Wright's fixation indices were used to analyze the level of population structuring. Twenty-nine LEI0258 allele sizes were observed using capillary electrophoresis.Allele sizes ranged from 185 to 569 bp with no significant difference in allele frequencies between populations (P < 0.01). Allele frequencies were in Hardy and Weinberg Equilibrium in all population except in improved Horro and Dara chicken (P < 0.05). Excluding the tandemly repeated motif, the number of monomorphic and polymorphic sites are 412 and 35, respectively. Number of point mutation and indels are 33 and 17, respectively. The number of R12 CTTTCCTTCTTT repeats ranged from 2 to 18, while R13 was found monomorphic with a single motif CTATGTCTTCTTT. Sequences relationships reveal two distinct groups of alleles. The high diversity at microsatellite LEI0258 at Ethiopian indigenous village chicken populations supports the importance of the MHC region in relation to the disease challenges diversity faced by smallholder poultry production within and across Ethiopian agro-ecologies. We recommend that breed improvement programs ensure the maintenance of this diversity by selecting breeding stock as diverse as possible at the LEI0258 locus. The part of the thesis reports the genomic diversity of Ethiopian indigenous chicken through discovery and characterization of 21 million SNPs (72% novel) from 27 indigenous chicken populations (n = 260 birds) using whole genome sequencing.In each population, around 10 to 12 million SNPs are present, of which, 40-47% are a heterozygote.The mean SNP density across all population is 21 (±5) per kb, although it is much lower on the sex chromosomes (W = 0.4, Z = 11). Principal component and admixture analyses suggest the presence of four ancestral gene pools across the populations. Over 46% of the SNPs are located within genes, of which exonic and intronic SNPs account for 1.59% and 43.94%, respectively; while 31% of the exonic SNPs are non-synonymous. A large proportion of SNPs has low CHAPTER 1. INTRODUCTIONIn the last 50 years, the world human population was increasing dramatically and its growth is projected to reach 9.73 billion by 2050 and 11.2 billion by 2100 which makes the greater confront for agriculture to meet the growing food demand (FAO, 2017). More than a billion people around the world live in extreme poverty. Moreover, as of 2009 there have been marked increases in hunger, and the Food and Agriculture Organization of the United Nations(FAO) estimated that 1.02 billion people are undernourished (FAO, 2010) and most of these people are found in sub-Saharan Africa. 30% or more of children under 5 years of age are malnourished in many parts of this region. Similarly, Cicia et al. (2003) indicated that the profound changes that have taken place in agriculture in the past 50 years in both developed and developing countries have caused severe and undesirable impacts on the environment.Particularly, Ethiopia is known to have an estimated human population of over 82.64 million, growing at a rate of 2.4% per annum at which 85% the population relies on agriculture (45% of GDP and 85% of export earnings) (CSA, 2012) including livestock production. Livestock production is an integral part of the country's agricultural production system and has a profound contribution to the national effort to ensure food self-sufficiency both at the national and household level (Dessie, 2011). In particular, chicken production is playing a fundamental role in its considerable potential to improve the livelihoods of rapidly increasing populations of Ethiopia since long time ago. Numerous indigenous chicken populations are reported to adapt to different circumstances of various regions of the country. These chickens are being kept mainly by smallholders under backyard poultry farming mainly as a livelihood source of income. Padhi et al. (2016) mentioned that commercial chicken production is world widely increasing because of its better efficiency, leaving indigenous breeds to hardly compete with. However, under normal circumstance, it is usually suggested that rare or endangered breeds are often highly adapted and their performances should be measured comparatively, within their own environmental conditions than comparing with other breeds under improved or modified conditions or under intensive management (Dessie, 2011). In other words, examining these breeds with respect to the products for which they were selected and valued in the conditions under which they evolved very meaningful. Azage et al. (2006) have in their remark stated that many occasions where growth rate, prolificacy, or milk production have been measured and used as criteria to point up the inferiority of purebred indigenous stock over that of exotic imported breeds or their crosses.Needless to say, indigenous breeds are known to possess occasionally unique alleles pertinent to their adaptation (diseases resistance and climate) and breeding goals (quality egg and meat) (Zanetti, 2009). Moreover, they have also enormous use for scientific, cultural and economic use that urges us to conserve and improve for the use of the future (Koehler-Rollefson, 2014). Many scholars also advocate that utilization of indigenous genotypes by improving their competitiveness under the socioeconomic circumstances of their production environments is one of the practical options to ensure the conservation of genetic diversity. Despite these beneficial roles, indigenous chicken genetic resources are not getting greater attention for further improvement in the country. For instance, regardless of breed type, FAO (2007) reported that chicken population and an average number of chickens per household in Ethiopia have declined by 64% and 37% over the last 50 and 20 years, respectively. In Ethiopia, the Growth and Transformation Plan (GTP) of the government, has put ambitious targets to double the chicken meat production from the total meat production by 2030 as part of a broader and integrated livestock productivity improvement program (MOFED, 2010). However, due to the limited fundamental information on the genetic diversity of nationally acquired chicken populations, the absence of well-managed conservation genetics programmes and the uncontrolled genetic introgression between indigenous as well as exotic breeds are seriously threatening the future of many indigenous chicken populations of the country.Genetic diversity is seriously suffering from genetic erosion in several fields of genetic resources especially in farm Animal genetic resources (Zanetti, 2009). This is especially true for the chicken industry where very few genotypes provide the breeding basis for the industrialized production (Romanov and Weigend, 2001). Weigend and Romanov (2002) stated that chicken genetic resources are considered to be one of the most endangered. In this regard, since the start of commercial chicken breeding in the middle of the 20 th century, chicken genetic diversity has become partitioned among relatively few highly specialized lines. As a consequence, many dualpurpose breeds, resulting from centuries of domestication and breeding, are now at the risk of being lost (Hillel et al., 2003).In a similar way, the issue of chicken genetic resources conservation is also a hotspot topic for Ethiopia, a country facing with major agricultural productivity challenges. For instance, according to IBC (2004), only limited research and development activities on phenotypic and genetic characterization have been conducted on chicken breeds of Ethiopia to conserve, reveal and exploit its genetic potential. As a result, little is known about the genetic diversity, structure, and degree of admixture within and between indigenous chicken populations in Ethiopia. Most breeds/populations may disappear before they have been recognized and exploited for breeding improvement goals. Specifically, in spite of the presence of various reports on the potential of indigenous chicken populations in Ethiopia, there are few studies, e.g. Hassen et al. (2009), Alemayehu et al. (2003), Mwacharo et al. (2007), Wragg et al. (2012) and Desta et al. (2014), which have reported the diversity of Ethiopian chicken using microsatellites, and SNP markers with no studies so far reporting the analysis of full genome sequences data.Hence, taking all this facts and figures, understanding the level of genetic diversity at genome level within and among chicken populations is of paramount importance in identifying potential chicken populations for genetic improvement programs.Research QuestionsThe following research questions are addressed in this study:• Is there any difference in genetic diversity among chicken populations of Ethiopia?• How is the diversity partitioned within and among chicken populations?• What is the genetic relationship between the target indigenous chicken populations?• What are the regions and genes under strong selection pressures in Improved Horro and other indigenous chicken populations of Ethiopia?The following major hypotheses are tested:• Due to the local preference of chicken genotypes among communities in Ethiopia, low genetic diversity within the population and high genetic variation between populations are expected.• Chicken populations will show geographic structuring primarily induced by rare longdistance dispersal, followed by mutation and limited gene flow.• Due to exchange at the local market, genetic distances between chicken populations will follow geographic distances.The principal objectives of this study are the following:The principal objectives of this study are the following:1. To analyse the genetic diversity and population structure of Ethiopian indigenous chicken populations.2. To analyse the genetic relationship between indigenous chicken populations.3. To determine the genomic regions under strong positive selection pressure in indigenous chicken populations.Chicken population makes up a large majority (around 30 billion) of avian breeds globally (Jensen, 2005). Chicken domestic genetic resources comprise a wide range of breeds and populations including native, also called indigenous, population, breeds and/or ecotypes, fancy breeds, middlelevel food producers, industrial stocks and specialized lines (Weigend and Romanov, 2001). Some archaeological and molecular evidences ascertain that the domesticated chicken (Gallus gallus)populations evolved in Asia from a common ancestor, the red jungle fowl Gallus gallus (Kanginakudru et al., 2008;Nguyen-Phuc and Berres, 2018;Sawai et al., 2010;Storey et al., 2012). Molecular and archaeological findings also support more multiple maternal origins (Liu et al., 2006) with possible contributions from other junglefowl species (Boichard et al., 2012;Eriksson et al., 2008;Gifford-Gonzalez and Hanotte, 2011;Mwacharo et al., 2013). For instance, the yellow skin locus present in several domestic chicken breeds is believed to most likely originate from the grey jungle fowl Gallus sonneratii (Eriksson et al., 2008).The time and the geographic center (s) of domestic chicken remain (Liu et al., 2006) uncertain.Some evidences confirm that the history of fowl domestication dates back to about 2000 B.C. (Kanginakudru et al., 2008;Sawai et al., 2010;Yap et al., 2010). But, more recent archaeological evidences showed that a much earlier domestication might have occurred around 6000 BC (Gifford-Gonzalez and Hanotte, 2011). Olusij (2010) indicated that the evolutionary history of the domestic fowl occurs in phases starting from the evolution of the genus Gallus, the emergence of the domestic fowl from its progenitors followed by the appearance of the current breeds, varieties, strains, and lines. During the course of domestication, the chicken has been considerably changed and diversified by natural and artificial selections (Al-Nasser et al., 2007). For instance, the ancestor of the domestic chicken, the red jungle fowl, lays 10 to 15 eggs per year in the wild, whereas commercial laying hens are capable of producing more than 300 eggs per year (Weigend and Romanov, 2001) following human selection and the global emergence of the chicken meat and egg industries (Siegel et al., 2006;Toro et al., 2014). Mwacharo et al. (2013) indicated that the history of introduction and dispersal of village chickens across the African continent is a subject of intense argument and speculation among scholars. Socio-cultural, linguistic, archaeological and historic data all suggest the introduction of chickens to Africa is through multiple maritime and/or terrestrial routes over time. Dessie (Alemayhu, 2003) also suggested that chicken could have been introduced into Africa through the Isthmus Suez, the Horn of Africa and through direct sea trading between Asia and Eastern Africa. So far, no molecular genetics study has attempted to understand the genetic diversity and origin of African indigenous chickens at the continental level (Gifford-Gonzalez and Hanotte, 2011).The poultry sector in Ethiopia can be characterized into three major production systems(largescale commercial, small-scale commercial and scavenging) based on some parameters such as breed, flock size, housing, feeding, health, technology and biosecurity (Tadesse, 2015). The most dominant type of poultry production system in Ethiopia is the scavenging production system inherently characterized by low productivity (Alemayhu, 2003;Halima et al., 2007). In this system, birds rely on what they are able to pick in the homestead with little or no supplementation with food leftover waste and a small amount of grains. This system is characterized by a low input, minimal level of biosecurity, high off-take rates and high levels of mortality (Desalew et al., 2013).Here, there are little or no inputs for housing, feeding or health care. It does not involve investments beyond the cost of the foundation stock, a few handfuls of local grains, and possibly simple night shades, mostly night time housing in the family dwellings. Over 80-97% are indigenous and raised in small flocks (4-10 hens) and produce a maximum of 40 eggs/hen/year and to achieve a market live-weight of less than 1.5 kg at 6 months (Dessie et al., 2011). Still, under this situation, indigenous chicken populations represent an important resource from which improved lines can also be developed. Here, evidences are emanating from well designed, selective breeding programs of indigenous populations showing that significant productivity improvement can be achieved, while maintaining a reasonable level of adaptation, even under scavenging conditions.For instance, this was manifested by the works on the improvement of Horro chicken populations in Debre Zeit Agricultural Research Centre (Dessie et al., 2011). Progress reports showed that survival has improved from less than 50% in the base generation to 98% in generation 6. Similarly, body weight at 16 weeks has increased substantially from 550g to 880g, while egg production per annum has got tripled from 24 to 72. These evidences pinpoint the potential of indigenous chicken populations for productivity improvement. However, these genetic resources are generally under exploited and under leveraged due to lack of effective capacity for local testing, multiplication, and delivery to farmers, followed by continuous genetic improvement. In fact, genetic improvements have been attempted by the government and several development agencies by rather crossbreeding programs of exotic commercial cocks with indigenous chickens than within breed improvement. Selection programs targeting to genetically improve, multiply and distribute improved indigenous chickens to the farmers are still largely to come. In Ethiopia, much of the chicken improvement efforts have focussed on delivering exotic and more often inappropriate birds to the smallholders' context. For instance, chicken genetic improvement programs in the previous decades focused on the use of White Leghorns, Rhode Island Red, Brown Leghorns, NewHampshire, Cornish, Australoup, Light Sussex and Fayoumi breeds all demanding high input and intensive management (Dana, 2011). These high-yielding breeds are not a sustainable option for improving village poultry so long as the production conditions are not suitable to such genotypes none adapted to the low nutrition input, high disease incidence and weather patterns of a country like Ethiopia. Many scholars advise that breeding programs should be oriented in such a way that it can address the underlying socioeconomic and production circumstances of village chicken production systems.The second poultry production system in Ethiopia is the small-scale commercial one. In this system, modest flock sizes usually ranging from 20 to 1000 exotic birds are kept for operating on a more commercial basis. Most small-scale poultry farms are located around Debre Zeit town in Oromia region and Addis Ababa. This production system is characterized by a medium level of feed, water and veterinary service inputs and minimal to low biosecurity.The third type of poultry production system is a large-scale commercial system. This intensive production system involves keeping to 10,000 birds or even more under indoor conditions with a medium to high biosecurity level (Desalew et al., 2013). This system heavily depends on imported exotic breeds that require intensive inputs such as feed, housing, health, and modern management system. It is estimated that this sector accounts for nearly 2% of the national poultry population in Ethiopia. This system is characterized by a higher level of productivity where poultry production is entirely market-oriented to meet the large poultry demand in major cities. The existence of somehow better biosecurity practices has reduced chick mortality rates to merely 5%. Key demographic and production parameters in poultry production systems are presented in Table 1. Indigenous chickens in Ethiopia are found in every corner of the country. They are not exhaustively but are closely related to the jungle fowl. Chicken populations in Ethiopia are non-descriptive which vary in plumage color, comb type, body weight, and body conformation. Under Ethiopian Chicken description context, no chicken was reported as a distinct breed despite some naming of distinct ecotypes. Living in different agro-ecologies they are often also referred as ecotypes. In most of the native flocks, broodiness (maternal instinct) is pronounced. They are characterized by slow growth, late maturity and low production performances (Desalew et al., 2013). The productivity of local scavenging chicken is low with high mortality of chicks. Duration of brooding time of a hen is wider with many cycles per year. Moreover, the low productivity of indigenous stocks partially is attributed to the low management standard of the traditional production systems. Provision of vaccination, improved feeding, clean water, and nighttime enclosure relatively improves the production performance of indigenous chickens (Habte et al., 2013).By and large, studies conducted on indigenous chickens to evaluate their performance usually ignore their unique physiological and behavioral characteristics and their socio-cultural values (Alemayhu, 2003). Due to their high genetic diversity, there is also remarkable variation in the performance of indigenous chickens within and among breeds. This variation is an important genetic attribute of the indigenous chicken, whereby selection can act to improve their performance (Ibid). About sixty million birds (n = 60,505,327) are available in Ethiopia for consumption, sale, breeding and socio-cultural values (CSA, 2016). Smallholder farmers keep birds for its low capital requirement, flexible production systems, and low production risk, taste, flavor and leanness (FAO, 2010;Halima et al., 2007). Under Ethiopian context, chickens are considered as the only type of livestock that many poor people can maintain relatively risk-free compared to another type of livestock enterprises (Dessie et al., 2011). It is providing a considerable potential to improve the livelihoods of many pro-poor communities of the country.However, low productive chicken genotypes dominate smallholder production systems, mainly owing to the absence of sound long-term chicken genetic improvement, multiplication and delivery systems (Dana, 2011;Padhi, 2016).Improving chicken productivity is the first step to get out of poverty by improving family nutrition and socio-economic status (Dessie et al., 2011). Numerous productivity evaluation elucidates that indigenous chickens have remarkable performance than their improved counterparts under low to medium input systems. Although indigenous chickens are known to have a number of adaptive traits in hot and humid tropics such as the necked neck, minimum and frizzle feathers, black bones and meats, the potential value of indigenous breeds remain under-estimated (Alemayhu, 2003;Dana, 2011;Fathi et al., 2017Fathi et al., , 2013)). Even though indigenous chickens are not fat growers and poor layers of small-sized eggs(about 45g), they are ideal mothers, good brooder, and excellent scavengers, sturdy and believed to possess better natural immunity against common poultry diseases (Alemayhu, 2003). Berthouly et al. (2008) advise the need to explore and properly manage indigenous chicken genetic resources as they represent both a heritage and a reservoir of genetic diversity. Chicken meat offers an attractive lower-carbon alternative to beef. However, chicken consumption is currently low in Ethiopia compared with other countries which are planned to boost to 30% by 2030 aimed to increase the chicken population by 70 million (CRGE, 2011).A combined analysis of a dataset of 28 phenotypic and 7 genotypic publications on Ethiopian indigenous chicken across socio-physical (religion, elevation, and agro-ecology) shows the different morphological and morphometric attributes of indigenous chicken (Table S 3, Table S 4,   Table S 5). Different factors may affect the production characteristic of indigenous chicken and their phenotypes. For instance, religion may be expected to influence on cultural trait preferences of smallholders in regards to chicken phenotypes. Similarly, elevation and agro-ecology as a proxy of the natural selective forces may have shaped the morphology of the birds at major agro-ecological zones. Geographic localization across Ethiopia of the various studies examined here is shown in Figure 1.In term of socio-cultural environment, most of the studies (n = 20) reports the Christian Orthodox faith as the dominant religion in the geographic region of the chicken population reported here, followed by the Muslim one (n = 13), with few studies examining chicken population in Protestant (n = 4) and traditional (n = 3) faith areas (Table S 3). Elevation wise, most of the studies concentrate (n = 25) in elevations ranging from 1800-2400 (Table S 4). A sizable number of (n = 15) phenotypic studies were also carried out in elevations ranging from 1800-2400 masl. There are no studies in elevations less 500 and above 3200 masl. The different studies have variably distributed across agro-ecologies which most of them concentrated in Tepid to cool moist mid highlands (n = 9) and Tepid to cool sub-humid mid highlands (n = 8) (Table S 5). Tepid to cool sub-moist mid highlands (n = 1) and Hot to warm per humid lowlands (n = 1) are least represented.Various studies reported plumage coloration and pattern, earlobe, shank and skin color variants in flocks of IC of Ethiopia but with inconsistency in color definition between them (Table S 6; Table S  7). Colour is an important feature for most living organisms, having great significance in the wild by affecting the survival and reproductive success of the species. For instance, carotenoidbased ornaments (skin or feathers) in wild birds are considered to be signal of an individual's nutritional status or health, reflecting its foraging efficiency or immune status and are therefore implied to affect sexual attractiveness (Blas et al., 2006;Castaneda et al., 2005). On top of its advantage in terms of genetic variation, a diversity of colors may serve as a camouflage from predator attack and in turn boost productivity where a free-range scavenging system is predominant. Apart from other methods, genetic variations in chicken can be described using traits based on pigmentation variants and comb varieties (Dana et al., 2010). Accordingly, different scholars consider color diversity of chicken as an important parameter in phenotypic diversity studies. Early color variants were mostly selected for utility reasons or religious practices (Sheppy, 2011). In Ethiopia, chicken plumage has greater economic significance and influences the local breeding strategies (Dana et al., 2010). Among others, the most critical color variant considered by farmers is plumage color which is given attention in local market preference and breeding objectives. To this end, a high diversity of Ethiopian IC plumage has been noted in the different phenotypic studies (Table S 10).The most dominant plumage color variants are red (39.26%), black (39.14%) and white with two or more mixture of the dominant colors reported by different scholars (n = 14; Table S 7). In terms of its relationship with religion, plumage coloration is not significantly different across the dominant religion category (P < 0.05; Table S 18). Its relationship was not also significantly to vary across elevation category for all plumages except brown coloration (P < 0.05). Apart from the red and black dominant plumages, a diverse spectrum of plumages in variable proportion are reported. Red plumage is the commonest in Tepid to cool sub-humid mid-highlands (SH2; 39.26%) followed by Hot to warm humid lowlands (H1; 24.76%) agro-ecological zones (Table S 28). In contrast, there is no report of red plumage chicken in cold to very cold sub-humid sub-afro alpine to afro-alpine (SH3) agro-ecology. Similarly, black plumage chicken reported predominating in SH2 (39.14%) and H1 (21.93%), respectively. Apart from the red and black dominant plumages, a diverse spectrum of plumages in variable proportion are reported. Only red, and zagolima plumages significantly vary across agro-ecologies (P < 0.05).Earlobe color is a naturally and artificially selected trait in chicken. In breeding operations, it has been selected as a breed characteristic as a head furnishing trait. White/red earlobe color is a polygenic and sex-linked trait (Luo et al., 2018;Nie et al., 2016). Variable earlobe color variants proportion are also reported from different studies in Ethiopia (Table S 11). Chickens with black (56.46%) and red (41.81%) earlobe variants are the commonest among other comb types. This variant does not vary significantly (P < 0.05) across dominant religion demography and elevation for all earlobe color types (Table S 20). Variable earlobe colour variants proportion are also reported across the agro-ecological zones. The highest number of chickens with red earlobe variants (41.81%) is found in (Tepid to cool sub moist mid highlands (SM2) agro-ecological zone (Table S 11). While, chicken with black earlobe variants are the commonest in Tepid to cool moist mid highlands (M2; 56.46%) and it reaches 40.14% in Tepid to cool sub-moist mid highlands (SM2). Unlike chicken with black earlobe colour, chicken with white earlobe colours are found across a wide spectrum of agro-ecological zones. Except for brownish and multi-colour variants, other reported colors vary significantly (P < 0.05) in terms of distribution across the respective agro-ecologic zones. Earlobe colour has no significant variation across agro-ecologies for indigenous chickens (P < 0.05, Table S 20) According to Jin et al. (2014), in chicken, skin pigmentation such as shank color is related to the levels of carotenoids and melanin. Yellow shanks are the commonest (50.98%) followed by white shank (25.59%) and black (10.59 %) shank color variants in Ethiopian IC (Table S 12). The rest of the reported shank color variants show the least occurrence across the different reported study sites. To this end, the least occurred shank color across studies is the mottled type (0.14%). Shank color is not significantly variable (P < 0.05) across dominant religion demography and elevation category (Table S 22). White shank color variants significantly vary across agro-ecologies unlike yellow and black shank color variants (P < 0.05, Table S 29). The presence of feathered shank gene (Pti/pti) in indigenous chicken is also minimal (Table S 17). A variety of skin variants are also documented in different spatiotemporal studies. In this respect, white (48.87%) and yellow (30.86%) skin color variants are the most abundant phenotype across various socio-geographic factors (Table S 12). In contrast, there is little black (0.35%) and green (0.20%) skin color variants in M2 agro-ecology. There was no significant difference in the skin color of indigenous chicken across dominant religion, elevation and agro-ecologies (P < 0.05; Table S 23; Table S 32).Like other coloration variants, eye color variants vary according to the carotenoid pigmentation and blood supply in the eye (Crawford, 1990). Under the context of Ethiopian IC, Eye color variants were only considered in few studies (n = 6, Table S 6; Table S 13). The most prominent eye color variants across these 6 studies are red (36.95%), black (23.80%), while, the least eye color variant is pearl (0.6%). The most prominent eye color variants across these 6 studies are red (36.95%), black (23.80%), while, the least eye color variant is pearl (0.6%). Eye color variants have no any relationship between the different dominant religion demographies and elevation categories (Table S 21). Unlike other plumage and earlobe color variants, eye color variants are only found in a few agro-ecological zones, with a single variant observed for hot to warm humid lowlands (H1) and hot to warm sub-moist lowlands (SM1) agro-ecological zones. Only blue black and dark brown eye color variants are shown to significantly vary across agro-ecologies (P < 0.05, feather color variant was also only reported in a couple of studies (n = 2; Table 6 with a predominance of brown (25.18%) followed by white color variants (20.73%). Neck feather variant types do not vary significantly (P < 0.05) across dominant religion demography. Neck feather has also no significant variation (P < 0.05) with elevation category except for white neck feather type coloration (Table S 26) while white (48.87%) and yellow (30.86%) skin color variants are most widespread across study sites. None of the neck feather color variants are reported to significantly vary across agro-ecologies (P < 0.05, Table S 31).Indigenous chicken (IC) feather genotypes reported in Ethiopia include normal feathered, crested head, frizzle, naked neck, and feathered shank (Table S 16). The naked neck (Na) gene is an autosomal, incompletely dominant gene described as one of the major genes in local chickens of the tropics that has desirable effects on heat tolerance and adult fitness (Fathi et al., 2013).However, to the scope of this review, little proportion of naked neck (2.75%) genes have been reported and distribute mainly in SH2 (66.57%) and H1 (24.78%) agro-ecology. This might be due to little availability of naked neck chicken (dominant genes) in Ethiopia or due to a single available study by Teketel (1986) as cited by Dana (2010) to exactly define the size and geographic distribution of these chicken. Moreover, this may be due to the fact that naked neck chicken is less preferred and have less aesthetic value to normal feathered and another type of chicken as they are considered ugly. Similarly, shank and feet (0.6%), as well as muff and beard (0.6%) feathered chicken, are available in extremely low frequency (Table S 16). About 97.73% of IC in Ethiopia reported having no shank feather. Another prominent phenotype in chicken is crest which is characterized by a tuft of elongated feathers atop the head. It shows an autosomal incompletely dominant mode of inheritance and is associated with a cerebral hernia (Wang et al., 2012). The crested head genotype is considered to be a superior egg producer (Ngeno et al., 2014). This gene is also rare in the flocks of Ethiopian IC (4.33%). Fewer proportion of crested birds have been reported and there is no work in birds with mutant phenotypes of the crest, necked nake, feathered shank, and frizzle adaptive genotypes in Ethiopia. The other genotype of Ethiopian IC chicken reported in various studies is frizzle (F) phenotype which is believed to be caused by a single autosomal incomplete dominant gene in which heterozygous individuals show less severe phenotypes than homozygous individuals (Ng et al., 2012). The rare abundance of frizzled (1.85 %) IC was also evidenced. The proportion of female chicken with a spur (43.65%) is also indicated in Table S 17. Normal feathered and crested chicken occurrence significantly vary across agroecologies (P < 0.05, Table S 24).Comb type is a trait that shows considerable variability among domestic chicken (Shen et al., 2016). This variability was also ascertained in Ethiopian IC chicken (Table S 15). The dominant comb types are single (42%), rose (31%) and pea (21%), while other comb types make the rest (7%). The remaining proportion include Walnut (2.42%), V-shape (2.18%), strawberry (0.34%), Cashion (0.38%), duplex (0.69%), buttercup (0.01%) and unclassified comb types (0.03%). Single comb type variants show a significant difference (P < 0.05) across dominant religion demography and elevation category, while, no variation is found for the rest comb type variants (Table S 19).The higher number of these comb type variants predominate in Tepid to cool sub-humid midhighlands (SH2) followed by Hot to warm humid lowlands (H1) agro-ecology (Table S 33). Only single, rose and duplex comb types significantly vary across agro-ecological zones in contrast to the consistent occurrence for the range of other comb types (P < 0.05).Body shape variants are presented in Table Table S 17 with a high predominance of blocky (59.43%) body shape type. Body shape does not significantly vary (P < 0.05) with the dominant religion demography except for blocky body shape (Table S 27).Body weight is an important attribute in poultry production as it forms the basis for not only assessing growth and feed efficiency but also making economic and management decisions (Dahloum et al., 2016). The body weight of IC chicken ranges from 0.84-1.97 Kg (Table 2). The annual egg production per hen is 47-75 with the lowest (47 ± 5.07) and highest (75 ± 5.07). On top of body weight and egg number, body length and shank length are intensively studied traits involving a higher number of chicken individuals and chicken populations. Various quantitative traits measurements have been also reported in previous studies of Ethiopian IC (Table 2). On top of body weight and egg number, the overall least significant means and their test of significance is indicated for shank length and other phenotypic traits (  Few genetic studies have been undertaken in IC chicken populations of Ethiopia (Table S 2).Among genetic diversity estimators He (n = 2), Ho (n = 8), MNA (n = 6), PIC (n = 3) and FIS (n = 5) are most studied in diversity studies in indigenous chicken populations of Ethiopia. However, considering the diversity and wide land size of the country, the studies are not representative in terms of sample size and physical coverage. Meta-analysis of results from combined studies indicate the range values for MNA (4.20-6.29), ENA (2.5-6.35), PA (2-3.75), He (0.12-0.63), allele/locus (4.7-5.10), Ho (0.27-0.93), FIS (0.02-0.17) and IBS (0.01-0.02) (Table 3). Only the effective number of alleles vary significantly across the different dominant religious demography (P < 0.05;   Over all, in the process of reviewing the different phenotypic and molecular studies throughout Lack of focus in targeting traits having significant socio-cultural and economical importance before designing of their diversity study was also another limitation of previous works. In this regard, the entire phenotypic studies only followed potentiality of chicken production.Besides, Ethiopian phenotypic studies lack to mention measurement devices and if all measurements were taken by the same individual to avoid individual induced systematic errors.Studies that run PCA and develop a prediction equation for body weight are not reported in this review. No phenotypic studies also attempt to determine the phenotypic diversity indices and quality of distribution of the traits using the Simpson's diversity index and quality of distribution.Dendo-grams through the squared Euclidean distance of hierarchal cluster analysis were not obtained in any kind of phenotypic studies in indigenous chicken populations of Ethiopia based on various physical factors.The chicken is the first sequenced animal agricultural species with the first of the chicken genome released in 2004 (Burt, 2007(Burt, , 2004;;Jensen, 2005;Tixier-Boichard et al., 2011). Chickens are being widely used as a model organism to study human diseases like muscular dystrophy, immunological diseases and thyroid insufficiency to find and investigate candidate genes affecting such traits (Boichard et al., 2012;Cogburn et al., 2007;Jensen, 2005). The chicken genome contains 15% of repetitive DNA sequences comprised of short tandem repeats and several families of long interspersed elements (Soattin et al., 2009;Treangen and Salzberg, 2012;Wicker et al., 2004). It comprises 39 pairs of chromosomes including eight pairs of macro-chromosomes, one pair of sex chromosomes (Z and W) and 30 pairs of microchromosomes (Burt, 2007;Rao et al., 2007). The size of the chicken genome (Gallus_gallus-5.0; GCA_000002315.3) is estimated to be 1.28 × 10 9(1,285,637,921) base pairs ( Table 4). The chicken genome contains 24,838 genes distributed over 39 pairs of chromosomes plus 34,811,469 LTR elements; 20,000 DNA transposons, 140,000 simple repeats, 571 ncRNA genes and 4,000 satellites (https://www.ensembl.org/Gallus_gallus/Info/Annotation). The chicken genome was also confirmed to possess about 2.8 million genetic polymorphisms (Fulton, 2009) and 5 SNPs per kb unlike the human genome (1 SNPs per Kb) (The 1000 Genomes Project Consortium et al., 2015). Simultaneously with the release of the 2004 chicken genome sequence, the Beijing Genome Institute identified and released the 2.8 million single nucleotide polymorphisms (SNP) to the public domain.Single nucleotide polymorphisms are single nucleotide variants within the DNA sequence. They were identified by comparing the Jungle Fowl genome sequence with partial sequence information (0.3×) from 4 different chickens: 1 Silkie (Chinese breed), 2 commercial broilers (meat type), and 1 inbred laboratory White Leghorn (egg-layer type). These SNP have formed the basis for all the large SNP genotyping platforms developed to date. The second build was released in 2006 that corrected some of the deficiencies found in the first version. However, even this second build had multiple deficiencies. It was missing many of the gene-rich micro chromosomes. Chromosome 16 contains the major histocompatibility complex (cluster of immune function genes) and was very poorly covered. The Z chromosome is known to be incomplete and to have a considerable number of errors in gene order (Fulton, 2012). A third build of the genome (Gallus_gallus-5.0; GCA_000002315.3) has brought an increase of N50 contig and scaffold size to 252 Kb (460%) and 12.4 Mb, respectively (Warren et al., 2016). Gallus Gallus 5.0 shows an increase of 4679 annotated genes (2768 noncoding and 1911 protein-coding) over those in Gallus Gallus 4.0 (Warren et al., 2016). A third build of the chicken genome has been produced but has not yet been released for public access.Table 4. Genome statistics for the chicken reference genome (Gallus Gallus 5.0) assembly. (http://www.ensembl.org/Gallus_gallus/Info/Annotation).Genetic diversity is a result of variations in DNA sequences among organisms due to mutations resulting from the substitution of single nucleotides (SNPs), insertion or deletion of DNA fragments, and duplication or inversion of DNA fragments (Fulton, 2009). It allows farmers to select stocks or develop new breeds in response to changing conditions to ensure food security (Toro et al., 2014). Besides, genetic diversity is important as an enormous number of livestock diversity is disappearing globally for a number of reasons to develop conservation and improvement strategies (Barker, 2001;Boettcher et al., 2010;Ruane, 1999;Simianer, 2005). A number of estimators of genetic diversity have been reported by different scholars.Allelic variability (number of alleles segregating in the population) is one of the estimators used to measure genetic diversity with key relevance in genetic conservation programmes (Barker, 2001;Simianer, 2005;Toro et al., 2014). The mean number of alleles (MNA) observed over a range of loci for different populations are a reasonable indicator of genetic variation provided that the populations are at mutational-drift equilibrium and that the sample size is almost equal for each population (Allendorf et al., 2010). Breeds with a low MNA have a low genetic variation which might be due to factors like-genetic isolation, historical population bottlenecks or founder effects and a high MNA implies great allelic diversity which could have been influenced by crossbreeding or admixture. The other estimators of allelic variability are the effective number of alleles (ENA) and allelic richness (Ar) (Allendorf et al., 2010). These parameters are used when the sample sizes are not the same for each population under subject. ENA denotes the number of equally frequent alleles it would take to achieve a given level of gene diversity. It allows to compare populations where the number and distribution of alleles differ drastically. Ar is a measure of the number of alleles per locus but allows comparisons to be made between samples of different sizes by using the rarefaction technique or a Bayesian simulation approach to standardize populations to a uniform sample size.The other fundamental parameter used to measure genetic diversity between and within populations is Hardy-Weinberg Equilibrium (HWE) (Zhou et al., 2009). A population is said to be in HWE when gene and genotype frequencies remain constant from generation to generation without the factors (e.g. selection, migration, and mutation) which can cause changes in these frequencies and in turn non-random union of gametes. Deviation from HWE in a population indicates possible inbreeding, population stratification and sometimes problems with the genotyping (Sha and Zhang, 2011). Allendorf et al. (2010) mentioned that the data required to test HWE in a natural population are gene and genotype frequencies and the size of the sample population at each locus. The χ 2 -test remains the most popular option (Salanti et al., 2005;Shriner, 2011) and tests of HWE are commonly performed using a simple x 2 goodness-of-fit test that compares expected and observed numbers of heterozygotes and homozygotes (Allendorf et al., 2010;Allendorf and Luikart, 2007;Wigginton et al., 2005).Another estimator of genetic variation is Linkage disequilibrium (LD) which is described as the non-random association between different loci which may arise from admixture of populations with different gene frequencies; chance in small populations (e.g. endangered breeds); selection favouring one combination of alleles over another; the close association between markers in the same linkage group (Lee et al., 2012). LD between densely spaced, polymorphic genetic markers in different species contains information about historical events of recombination in a population (Hayes, 2003;Khanyile, 2015). Most measures of LD, such as r 2 and related measures, quantify the association between a pair of loci. The degree of linkage disequilibrium can be estimated directly from the genotypic frequencies in a sample of individuals taken from the population (Hayes, 2003).The average expected heterozygosity (He) also called Nei's gene diversity; defined by Nei (1986) at n loci within a population; is the best general measure of genetic diversity within a population (Allendorf and Luikart, 2007). Individual breed average heterozygosity is estimated by summing heterozygosity at each locus and averaging these values over all loci (Hedrick and Kalinowski, 2000). It is often calculated based on the square root of the frequency of the null (recessive) allele as follows: He=1-∑ p n i i 2 , where pi is the frequency of the i th allele (Allendorf et al., 2010;Allendorf and Luikart, 2007). The observed heterozygosity (HO) is defined as the percentage of loci heterozygous per individual or the number of individuals heterozygous per locus. High heterozygosity values for a breed may be due to long-term natural selection for adaptation, to the mixed nature of the breeds or to historic mixing of strains of different populations. A low level of heterozygosity may be due to isolation with the subsequent loss of unexploited genetic potential.In genetics, inbreeding is defined as the mating of closest relatives (Allendorf and Luikart, 2007).The inbreeding coefficient (f) is the probability that the two alleles at a locus within an individual are identical by descent as they are derived from the same allele in a common ancestor in a previous generation. It is recommended that inbreeding coefficients should only be estimated for breeds which show significant deviation from the HWE (Curie- Cohen, 1981;Vieira et al., 2013). A large value reflects the existence of a small number of heterozygote genotypes and an excess of homozygote genotypes. A small value indicates the occurrence of heterozygote genotypes at a higher proportion than the homozygote genotypes.When a population is divided into subpopulations, there is less heterozygosity than there would be if the population was undivided (Kanginakudru et al., 2008). There are different types of approaches enumerated to quantify the distribution of genetic diversity within and between livestock populations. But, the two commonly used approaches to quantify the distribution of genetic diversity within and between populations include : Wright's F statistics (fixation indices) and AMOVA (Analysis of molecular variance). Computations of Wright's fixation indices (FIT, FST, and FIs) are pivotal and most widely used for studying the genetic differentiation of populations (Nei, 1986). The fixation index ranges from 0 (indicating no differentiation between the overall population and its subpopulations) to a theoretical maximum of 1. In other words, 0 indicates identical allele frequencies in a pair of populations (no differentiation) and 1.0 indicates alternate fixation for a single unique allele (the absence of any shared alleles) in each population.In practice, however, the observed fixation index is much less than 1 even in highly differentiated populations (Bird et al., 2011). The most commonly used programs for performing AMOVA are Arlequin, GDA, and GenAlEx (Excoffier and Smouse, 1992).Phylogenetic analysis is the means of inferring or estimating evolutionary relationships among breeds or populations and categorize cattle populations (Brinkman and Leipe, 2002). The evolutionary history inferred from the phylogenetic analysis is usually depicted as branching, treelike diagrams that represent an estimated pedigree of the inherited relationships among molecules, organisms, or both. The commonly used methods of clustering fall into two general categories: hierarchical and non-hierarchical. Hierarchical procedures are the most commonly used in animal diversity studies called phylogenetic analysis. The genetic distance measures are used to construct the dendrograms, also called phylogenetic trees. The two most commonly used methods for constructing the trees are unweighted pair group method (UPGMA) and the neighbor-joining method (NJ) (Backeljau et al., 1996). UPGMA trees give an indication of the time of separation (divergence) of breeds. The higher the branch length the longer is the separation period between breeds (Brinkman and Leipe, 2002).Molecular (genetic) markers are defined as any stable and inherited variation that can be detected by a suitable technique to subsequently detect the presence of a specific genotype or phenotype other than itself (Fulton, 2009). The development of these markers has been created new opportunities for the selection and genetic improvement of livestock. Practically, DNA based polymorphisms are being used for marker-assisted selection strategies, parentage testing, species identification, and population genetic studies (Naqvi, 2007). These polymorphisms also provide the foundation for genetic linkage maps, which are being used to identify loci for economically important variation and speed up the rate of improvement in production traits (Zhang et al., 2015).The use of molecular and bio-chemical markers to predict the total genetic merit of livestock to redesign animal breeding and management programs were also appraised (Toro et al., 2009).Molecular markers are playing a role in estimating the diversity (Bruford et al., 2003;Gibson et al., 2006;Teneva, 2009;Weigend and Romanov, 2002), distinctiveness, population structure of animals (Gholizadeh and Mianji, 2007) and aid in the genetic management of small populations, to avoid excessive inbreeding (Zanetti, 2009). Among others, the most pronounced DNA-based molecular techniques that are used to evaluate DNA polymorphism in chickens improvement programs are reviewed as follows:Restriction Fragment Length Polymorphism (RFLP) is the most broadly used hybridization-based molecular marker (Tazeb, 2018). RFLPs are inherited as naturally occurring Mendelian characters and have their DNA rearrangements due to evolutionary processes, point mutations within the restriction enzyme recognition site, mutations within the fragments, and unequal crossing over.The advantage of RFLPs is that they are co-dominant markers and are very reliable in linkage analysis and breeding. The limitations of the RFLP marker are that a large amount of DNA is required for restriction digestion and Southern blotting. The RFLP is relatively expensive and hazardous due to the requirement of a radioactive isotope. The assay is time-consuming and laborintensive and only one out of several markers may be polymorphic, which is highly inconvenient especially for crosses between closely-related species which precluded its widespread adoption within the poultry breeding industry (Fulton, 2012). Their inability to detect single base changes restricts their use in detecting point mutations occurring within the regions at which they are detecting polymorphism.Another method for detecting polymorphic markers is the random amplified polymorphic DNA (RAPD) assay. This assay, which is based on the polymerase chain reaction (PCR), uses short oligonucleotide primers of arbitrary sequence to amplify discrete regions of the genome (Kumar and Gurusubramanian, 2011;Tingey and de, 1993). This marker was developed by William and his co-workers in 1990, which employs single primer usually with 10 nucleotide bases as oligonucleotide primers and a GC content of at least 50% to amplify discrete fragments of DNA in low stringency of polymerase chain reaction (Smiths et al., 1996). RAPD is a quick, rapid, and inexpensive method of studying the DNA polymorphisms within and between populations based on the amplification of random DNA segments with single primers of an arbitrary nucleotide sequence. This dominant marker has also been widely used in poultry research (Shivashankar, 2014). The RAPD markers were used to detect polymorphism among five breeds of chicken i.e.White Leghorn and Rhodes Island Red (selected for part period egg production and egg mass respectively), Red Cornish and White Plymouth Rock (selected for early body weights) and Kadaknath (native breed) (Sharma et al., 2001). It was also used to evaluate genetic diversity and relatedness within and among four breeds of chickens and two turkey populations (Smiths et al., 1996). Among the PCR based DNA marker, RAPDs are cost-effective, most versatile and relatively easy to perform. The technique requires no prior knowledge of DNA sequence and utilizes minor quantities of DNA material, therefore, can be applied to even rare plant species. The main limitations encountered with the use of RAPD markers are repeatability of banding patterns and dominant inheritance. The ambiguity of the resulting fingerprint patterns and the fact that heterozygotes cannot be distinguished from homozygotes due to its dominant inheritance mechanism is another limitation of RAPD (Rege and Okeyo, 2006). In addition, how the genetic variation observed is generated is not fully understood, making reconstruction of evolutionary histories from RAPD data difficult.Microsatellites are also known as simple sequence repeats SSRs (Bruford et al., 2003); short tandem repeats (STRs) or simple sequence length polymorphisms SSLPs are the smallest class of simple repetitive DNA sequences. They are co-dominant, highly polymorphic, and multi-allelic.They are well-dispersed throughout the genome and are presumed to be selectively neutral.Microsatellite polymorphism refers to the differences in allele sizes due to variation in the number of repeats of base sequences that are detected by gel electrophoresis (Rege et al., 2006). They are known to be universal in prokaryote and eukaryote genomes and are present both in coding and non-coding regions. Genetic diversity measures using the highly polymorphic variable number of tandem repeat loci yield reliable and accurate information for the study of genetic relationships between chicken populations (Weigend and Romanov, 2002). In addition, microsatellites are easy to identify and have low mutation rates (Zhang et al., 2002). Many microsatellites have recently become available and being exploited in chickens, and have been mapped in reference (Gholizadeh and Mianji, 2007). Based upon sites in which the same short sequence is repeated multiple times, they present a high mutation rate and codominant nature, making them appropriate for the study of both within-and between-breed genetic diversity (Toro et al., 2009).Amplified Fragment Length Polymorphism (AFLP) are based on the detection of restriction fragments by PCR amplification. Genomic DNA is restricted with two different restriction endonucleases and then a subset of these are amplified using a modified PCR and visualized using radioactivity, silver staining or fluorescent dyes for use with an automated sequencer (Rege and Okeyo, 2006). Advantages of the AFLP are that no prior sequence information of the genome is required. A large number of polymorphic bands are produced and the technique is highly reproducible and standardized kits are available (Duim et al., 2000). According to the narrations of Toro et al. (2009), the AFLPs are dominant bi-allelic markers that provide an easy way to carry out a genome-wide screening of variation. AFLPs are reliable informative multi-locus probes and provide high levels of resolution that allows delineation of complex genetic structures (Rege and Okeyo, 2006). They have the disadvantage of a reduced power to analyze within-breed diversity due to the dominant mode of inheritance but could be very useful in analyzing between-breed variation.Single nucleotide polymorphism (SNP) is a new and very promising molecular marker system which offers opportunities to assess the genetic diversity in farm animal species differently by investigating the mode and extent of changes in specific positions in the genome (Weigend and Romanov, 2002). They can be explained as any polymorphism between two genomes that is based on a single nucleotide exchange (Fulton, 2008). These are the most frequent type of variations found in DNA and their discovery together with insertions/deletions has formed the basis of most differences between alleles (Gu, 2004;Ye et al., 2007). SNPs may occur in the coding, non-coding (most frequent in this region) and intergenic regions of the genome. SNPs represent one of the more interesting approaches in animal identification because they are abundant in the genome, genetically stable and amenable to high-throughput automated analysis (Teneva and Petrovic, 2010). On average, SNPs occur every 1000 -2000 bases and thus could be used as a genetic marker to systematically explore SNP variants for associations with quantitative traits (Gu, 2004;Ye et al., 2007). In the chicken genome, SNP has been identified with a frequency of 1 SNP per 225 bp, which is 5 times as many as in humans (Jalving et al., 2004;Orsini et al., 2011). Toro et al. (2009) have indicated their potential to detect both neutral and functional genetic variation because, although most of them are located in non-coding regions, some correspond to mutations inducing changes in expressed genes.In animal breeding and genetics, SNP offers opportunities to assess the genetic diversity in farm animal species differently by investigating the mode and extent of changes in certain positions in the genome (Weigend and Romanov, 2002) and for association mapping of genes controlling complex traits and provide the highest map resolution. SNPs offer several advantages over other types of DNA marker systems and are rapidly becoming the markers of choice for many applications in genome analysis due to their abundance (especially important in linkage disequilibrium-based mapping approaches) and also because high throughput genotyping methods are being developed for their analysis. The additional advantage offered by this approach lies in the phylogenetic information gathered through sequence variation analysis that allows drawing inferences on allele and population history that cannot be gathered with any of the other marker systems available. SNPs are also evolutionarily stable (i.e. do not change much from generation to generation) making them easier to follow in population studies (Rege and Okeyo, 2006). Groenen et al. (2011) have affirmed that a high success rate of the SNPs on the Illumina chicken 60k bead chip. Johansson and Nelson, (2015) have used the 60k SNPs to characterize the genetic diversity and map loci associated with a trait that are segregating in both Swedish local chicken breeds. Kranis et al. (2013) also developed a very high-density 600k SNP genotyping array for chicken commercial lines. With regard to the mapping of genes of interest in indigenous chickens, so far a single work has been done by Wragg et al. (2012) who used 15 randomly selected 5 chicken Ethiopian populations (in Gondar, Konso, Gumuz, Sheka, and Guduro) together with other breeds.They mapped phenotypic traits (skin and egg color) using genome-wide association study and revealed the possibility of fine mapping of non-pedigree chicken populations to characterize Mendelian traits at the molecular level using association analysis. Wragg and his colleagues also recommended that at least 90-110 kb SNPs are required for effective genome-wide associations study in village chickens. Their study considered a few numbers of samples (three chickens from each population) and it used a 50 K SNP chips.Mitochondria are sub-cellular organelles containing an extra chromosomal genome that is separate and distinct from the nuclear genome. Mitochondrial DNA is maternally inherited, does not undergo recombination, is a valuable molecule in investigating phylogenetic relationships among populations, subspecies, and species, and can be used to evaluate the maternal genetic constitution for a specific population (Shen et al., 2002). Shen and his colleagues in their study have made the first attempt to estimate the level of mtDNA diversity in the Chunky broiler and to compare it with some other chicken breeds used for egg production. They have presented new molecular evidence from the entire mitochondrial cytochrome b gene for the Chunky broiler and three other eggchicken breeds.The Major Histocompatibility Complex (MHC) is a highly conserved gene comprising a cluster of over 80 genes (92 kb) spanning on chromosome 16 (Chazara et al., 2011) characterized by high polymorphism and a tight linkage into a single supergene complex (Baelmans et al., 2005;Jarosinski et al., 2010;Miller et al., 2004), which have sought the attention of researchers for their association with disease resistance or susceptibility (Banat et al., 2013;Chazara et al., 2013Chazara et al., , 2011;;Gao et al., 2015;Hoque et al., 2011;Owen et al., 2008;Walker et al., 2011;Weigend and Romanov, 2001). Particularly, the chicken MHC on GGA16 has long been known as a gene region contributing significantly genetic resistance to infectious diseases (Hosomichi et al., 2010(Hosomichi et al., , 2009;;Miller and Taylor, 2016;Warren et al., 2017). It comprising loci encoding receptors which bind amino acid fragments from foreign pathogens on the surfaces of various immune and non-immune cells (Baelmans et al., 2005;Chen et al., 2012;Fulton et al., 2016;Jarosinski et al., 2010;Ncube et al., 2014;Nguyen-Phuc et al., 2016;Nikbakht and Esmailnejad, 2015;Zheng et al., 1999).LEI0258 is a highly polymorphic microsatellite locus located within the BF region of MHC-B on chromosome 16 and has been reported to have a direct association with chicken performance (Chazara et al., 2011;Figure  2). These include antibody responses to vaccination against Newcastle Disease Virus (NDV), Marek's disease, corona virus and coccidiosis (Baelmans et al., 2005;Briles et al., 1977;Nikbakht and Esmailnejad, 2015;Hateren et al., 2013). Its association with body weight, survival, embryonic mortality, fertilization rate, hatchability, egg production and resistance to worms has also been documented in many studies (Wang et al., 2014). Given the high level of polymorphism, LEI0258 marker genotypes have been suggested as good indicator of MHC-B haplotypes and it has become an important genetic marker used in chicken breed improvement program (Nguyen-Phuc et al., 2016). This has been confirmed by Chazara et al. (2013) who have ascertained that the LEI0258 marker genotypes an excellent predictor of the heterozygosity at the MHC locus. LEI0258 is described as atypical variable number tandem repeat (VNTR) which is composed of 12 bp (CTTTCCTTCTTT) and 13 bp (CTATGTCTTCTTT) conserved sequences which are flanked on both sides by indels and SNPs (Fulton et al., 2006). In an association study of MHC haplotypes with Marek's disease, Cole found 96.5% resistant of the birds with B-21 haplotypes were resistant to viral infection while of the birds with the B-19 haplotypes suffered 100% incidence of mortality. Polymorphisms are high at LEI0258. Fulton et al. (2006) observed allele size diversity ranging from 182 bp -552 bp. Lwelamira et al. (2008) genotyped two chicken ecotypes from Tanzania and they identified 22 and 23 alleles at LEI0258, respectively. They further report that allele 206 bp had a significant positive correlation (P < 0.001) with the elevated antibody responses against NDV vaccine, whereas the allele 307 bp was positively correlated with body weight trait.  (Izadi et al., 2011).The identification of regions that have undergone selection is one of the principal goals of theoretical and applied and evolutionary genetics (Gouveia et al., 2014). Such studies can also provide information about the evolutionary processes involved in shaping genomes, as well as physical and functional information about genes/genomic regions (Ibid). Artificial selection is the primary factor in the domestication and breeding history of livestock species. In the genomic era, selection refers to any nonrandom, differential propagation of an allele as a consequence of its phenotypic effect (Vitti et al., 2013). Selection may act in a directional manner, in which an allele is favored and so propagated (positive selection) or disfavored (negative selection, also called purifying selection) (Ibid). Positive selection leaves a more conspicuous footprint on the genome that can be detected using a number of different approaches (Jacobs et al., 2016;Pavlidis and Alachiotis, 2017;Ronen et al., n.d.;Wollstein and Stephan, 2015). The detection of \"signature of selection\" is now possible on a genome-wide scale in many plant and animal species and can be performed in a population-specific manner due to a wealth of per population genotype data that is available (Cadzow et al., 2014). Identification of the genomic signatures of recent selection may help uncover casual polymorphisms controlling traits relevant to recent decades of selective breeding in livestock (Fu et al., 2016) and can contribute to further shaping economically important traits (Ma et al., 2018).The tools used to detect evidence of selection are dependent on the nature of selective signature being investigated, which itself depends on the time scale over which selection occurred (Fu et al., 2016;Roosen et al., 2016;Tarekegn, 2016;Williams, n.d.). The Fst statistic has been a popular choice for investigation selection by utilizing differences in allelic frequency between populations to infer selective pressure in one population relative to the other and allowing detection of potential selection occurring in the range of 50,000 to 75,000 years prior for human populations (Ibid). A locus that shows significantly highest Fst statistics compared with other loci provides evidence of positive selection (Nielsen, 2005). Tajima's D is also another method suitable for detecting evidence of positive selection in human populations occurring within the past 250,000 years or approximately 10,000 generations and operates by identifying an excess of low to low intermediate frequency variants. Another commonly used measure is Fay and Wu's H (Fay and Wu,2000) which is useful for detecting evidence of more recent positive selection (80,000 years or approximately 3000 generations), particularly for intermediate-high frequency variants and thus complements Tajima's D and other methods (Baxevanis and Ouellette, 2001). Analysis of haplotypes provides another mechanism for identifying evidence of selection, with a number of methods utilizing the Extended Haplotype Homozygosity (EHH) concept. One of the more popular of these approaches is the Integrated haplotype Homozygosity Score (iHS) methodology, which provides a standard measure of the decay in EHH around a point (e.g., a SNP) from derived allele relative to the central allele (Voight et al., 2006). Regions of slowly decaying haplotype Homozygosity in the derived allele (longer than expected haplotypes, relative to the ancestral allele) are thus indicative of selection at that locus (Cadzow et al., 2014). Rubin et al. (2012), andElferink et al., (2012) have investigated selection signatures in a large number of chicken breeds using Z-transformed pooled Heterozygosity (ZHp) scores. This statistic estimates local heterozygosity depression in chromosomal regions and has been appropriately applied for detecting alleles that have swept to fixation or near fixation for long-term directional selection or during domestication (Fu et al., 2016;Rubin et al., 2010). Both the Fst statistic and Tajima's D can be calculated using standard genotype data obtained from heterozygous populations must be phased prior to calculation of iHS (Cadzow et al., 2014).  Equilibrium in all population except in Dara chicken (P > 0.05). Excluding the tandemly repeated motif, we identified 412 monomorphic and 35 polymorphic sites. The number of point mutationHigh evolutionary pressures occurred in chicken during the course of domestication and subsequent natural and human selection (Downing et al., 2009). Among others, infectious diseases exert strong selective pressures by affecting genes associated with innate and adaptive disease resistance and susceptibility. According to Salomonson et al. (2014), many of the genes involved in immunity are part of multigene families. In some families, each gene is conserved for a specific function dedicated to a particular outcome, in others allelic polymorphism and copy number variation allow rapid evolution in response to new environmental challenges, and other families comprise both kinds of genes (Miller et al., 2004). The Major Histocompatibility Complex (MHC) is one of these multigene family comprising loci encoding receptors which bind amino acid fragments from foreign pathogens on the surfaces of various immune and non-immune cells (Baelmans et al., 2005;Chen et al., 2012;Fulton et al., 2006;Fulton et al., 2016;Jarosinski et al., 2010;Ncube et al., 2014;Nguyen-Phuc et al., 2016;Nikbakht et al., 2013). MHC is a cluster of over 80 genes (92 kb) spanning chromosome 16 (Chazara et al., 2013(Chazara et al., , 2011;;Lima-Rosa et al., 2005;Miller et al., 2004;Nikbakht and Esmailnejad, 2015;Walker et al., 2011) characterized by high polymorphism and a tight linkage into a single supergene complex (Miller et al., 2004), which have sought the attention of researchers for their association with disease resistance or susceptibility.LEI0258 is a highly polymorphic microsatellite locus located within the BF region of MHC-B on chromosome 16 (Kannak et al., 2017;Miller et al., 2004). It has been reported to have a direct association with chicken performance and diseases tolerance (Nikbakht and Esmailnejad, 2015).It includes allelic variation in antibody responses to vaccination against Newcastle Disease Virus (NDV) (Baelmans et al., 2005;Nikbakht et al., 2013), Marek's disease (Fulton et al., 2016;Wang et al., 2014), corona virus (Hateren et al., 2013) and coccidiosis. Also, its association with body weight, survival, embryonic mortality, fertilization rate, hatchability, egg production and resistance to worms has also been documented in many studies (Owen et al., 2008). Given its high level of polymorphism and linkage disequilibrium association with the MHC-B locus, LEI0258 marker genotypes has been suggested indicator of MHC-B haplotypes, and it has become an important genetic marker used in chicken breed improvement programs (Banat, 2013;Gao et al., 2015;Hoque et al., 2011;Weigend et al., 2001). This has been confirmed by Chazara (2013) who have ascertained that the LEI0258 marker genotypes an excellent predictor of the heterozygosity at the MHC loci.LEI0258 is described as an atypical variable number tandem repeat (VNTR) locus which is composed of 12 bp (CTTTCCTTCTTT) and 13 bp (CTATGTCTTCTTT) conserved repeat sequences which are flanked on both sides by indels and SNPs (Fulton et al., 2006). In an association study of MHC haplotypes with Marek's disease, Bumstead (1998)  Indigenous chicken (IC) (Gallus gallus domesticus) are widely distributed in the diverse agroecological zones of Ethiopia. Accordingly, they represent ecotypes which may possess unique combinations of alleles in a given gene (Ngeno et al., 2015). Relatively, few works have been done so far on the genetic characterization at molecular level Ethiopian indigenous chicken. In particular, no studies have attempted so far to characterize the immune system of Ethiopian chicken. We report here the characterization and diversity of the MHC-linked LEI0258 microsatellite marker in 236 indigenous chickens from 24 distinct populations sampled across Ethiopia.Blood samples were collected from 24 chicken populations in Ethiopia (Figure 3). Samples included 80 cocks and 156 hens. Except for Tsion Teguaz and Meseret populations, two villages per population were sampled (8-10 chicken from each village). One or two chicken were sampled per household. Photographs and weight of each bird were taken. The average weight of sampled chicken was 1.26 Kg with age ranges of 5 to 36 months. Sampling included chicken from different agro-ecological zones with an altitudes ranges of 730 -3500 meters (Table 6). From the wing vein of each chicken, 50 -250 µl of whole blood was drawn with syringes using cryo-tubes filled with 1.5 ml absolute ethanol (100%) following the guidelines available at https://www.sheffield.ac.uk/nbaf-s/protocols_list.  Total DNA was extracted from chicken whole blood at the BecA-ILRI Hub, Nairobi, Kenya facility (http://hub.africabiosciences.org/) using the Qiagen DNeasy blood and tissue kit protocol (Lwelamira et al., 2008). To evaluate the DNA concentration a Thermo Scientific NanoDrop spectrophotometer 2000c was used. The integrity of DNA was confirmed by agarose gel electrophoresis whereby 20 ng/µl genomic DNA samples were loaded with 1 µl loading dye (6X) on 1% agarose gel containing 2.5 µl gel red at a voltage of 7/cm for 60 minutes, 3 µl of lambda DNA of size of 48,500 bp and at concentration of 20 ng/µl was used as size marker and the gel was then examined using UV light using a GelDoc-It 2 Imager to check the DNA quality and quantity. The total amount of DNA was normalized to 20 ng/µl using milliQ water for polymerase chain reaction (PCR) and genotyping. PCR amplification was carried out using a thermo-cycler PCR machine ABI PCR 9700 (Applied Biosystems). The primer sequences (GenBank accession number Z83781) for PCR amplification of LEI0258 were: forward 5'-CACGCAGCAGAACTTGGTAAGG -3' (length = 22 bp; GC content 47.6%; Tm = 71.5 C) and reverse-5'-AGCTGTGCTCAGTCCTCAGTGC-3' (length = 22 bp; GC content 46.2%; annealing temperature 69.9 C). The optimal PCR conditions were as described in (Gupta et al., n.d.;Han et al., 2013;Izadi et al., 2011;Nikbakht et al., 2013)  A 1 Kb ladder DNA from Bioneer was used as a reference to the size of the amplicons (Figure 5).The gel was exposed to UV light using GelDoc-It 2 Imager, to reveal the amplified fragments and their size. The forward primer sequence was labeled with the ABI fluorescent dye PET. electrophoresis were employed except that they were now tailed with T7 (20 base pair) and SP6(17 bp) tail sequences for the forward and reverse primers, respectively. Homozygote DNA fragments were purified using GeneJet PCR purification Kit (Thermo Fisher Scientific; cat. No.) and heterozygote DNA fragments were purified using the Qiagen Gel Extraction Kit and sent for sequencing to BIONEER sequencing platform in Korea. Alleles were sequenced on an ABI 3730XL DNA Analyzer using T-7 and SP6 sequencing primers.The genotypic data were subjected to various within and among populations genetic diversity analysis. These included: calculation of the total number of alleles, allelic frequency, and their distribution among the entire populations, polymorphic information content (PIC) for each population, Shannon's Information Index using GenAlEx software package version 6.5. The same package (Peakall and Smouse, 2012) was used to compute abundance of alleles (number of rare alleles and common alleles), partitioning of total genetic variation into within and among pre- expected (He) heterozygosity was estimated using the formula: \uD835\uDC3B\uD835\uDC52 = 1 −∑ (p 2 i +q 2 i) where p is the allelic frequency of the allele one at a given locus and q is the frequency of the alternate allele at the same locus. The deviation of each population from Hardy-Weinberg Equilibrium (HWE) was also tested using GenAlEx software package. Correlation between sampling site geographic distances and genetic distances between indigenous chicken populations was also done using the Mantel test (XLSTAT, 2018). Number of homozygote and heterozygote genotypes were calculated using power marker analysis (Liu et al., 2002).Population differentiation and the genetic structure were examined by sampling sites and Major Agro-Ecological Zones (MAEZ) on the basis of available allelic frequency using principal component analysis (PCA) calculated with the XLSTAT software (XLSTAT, 2018). Pairwise Fst among all pairs of populations were computed using the Weir and Cockerham statistics (Peakall and Smouse, 2012). Analysis of the population structure was done using STRUCTURE software version 2.3.4 using an initial length of 50,000 burn-in periods followed by 150,000 MCMC (Markov Chain Monte Carlo). Individuals were grouped into a predefined number of population clusters (K) ranging from 2 to 10. For each value of K, 20 independent runs were performed using the admixture and allelic frequency correlated models. Pairwise comparisons of the 20 runs were carried using the algorithm implemented in CLUMPAK server (CLUstering Markov Packager Across K) (http://hpc.ilri.cgiar.org/beca/bioinfo/clc.html). The best K was calculated following the equation proposed by Evanno et al. (2005).In addition, high-quality sequence reads with base call accuracy higher than 95% were assembled and resolved for conflicts using Qiagen's CLC work bench version 7. The resulting consensus sequences of chicken populations were aligned using the ClustalW program integrated into the MEGA (Molecular Evolutionary Genetics Analysis) software version 7 (Kumar et al., 2016). For this, a reference homologous sequences of the LEI0258 marker with SNP position was downloaded from the National Centre for Biotechnology Information (NCBI). The positioning of SNPs and Indels were done using the DnaSP 4.0 software package (Librado and Rozas, 2009). To determine the haplotype relationship of 13 new and 8 already NCBI described allele sequences, the median-joining network was done using Network 5.0.0.3. Based on the network only 10 haplotypes were identified and a tree was inferred by using the Maximum Likelihood method based on the Tamura-Nei model. Evolutionary analyses were conducted in MEGA7 (Kumar et al., 2016).    The genetic structure of all populations were examined by sampling sites and MAEZ on the basis of all available allelic frequencies using principal component analysis (PCA)(XLSTAT, 2018;Figure 11). The first two components axes accounted for 21.54% and 17.53% of the variation, respectively, for sampling site and 35.66% and 23.60%, respectively for MAEZ. Accordingly, the PCA of allelic frequency covariates by sampling sites, shows two distinct groups of the indigenous Ethiopian chicken populations separated longitudinally (Figure 6). This grouping seems to follow the direction of the rift valley geographically. Within these two groups, the chicken populations did not cluster in relation to their geographic origins. The PCA by MAEZ does not show any distinct genetic cluster for the indigenous chicken populations of Ethiopia (Figure 7). According to Evanno et al.(2005), the best clustering of the 24 chicken populations was found at K = 2 (Figure 12) both at predetermined population and agro-ecological zone level (Figure 13).The highest mean genetic distance (d = 0.94) was between Alfa Midir and Batambe followed by the former and Gafera (d = 0.90) populations (Table S 40;Table S 44). The average Fst for the entire population was 0.028. The pairwise Fst calculation shows the highest value (Fst = 0.12) between Hugub and Surta populations. In terms of analysis of genetic distance by MAEZ, the highest genetic distance is observed between Tepid to cool semi-arid mid highlands and hot to warm arid lowland plains (Table S 41). The Analysis of Molecular Variance shows the presence of 89% within individual variation across populations. The among-population variation and among individual variation was reported as 3% and 8%, respectively. Similarly, alleles, in tepid to cool sub-moist mid highlands are genetically distinct from alleles in Tepid to cool moist mid highlands (M2), Cold to very cold moist sub-afro alpine to afro-alpine (M3), Hot to warm sub-humid lowlands (SH1), Hot to warm sub-humid lowlands (SH2) agro-ecologies. The fixation coefficient of the subpopulation within the total population (FST), inbreeding/fixation/ coefficient of an individual in a subpopulation (FIS) and total inbreeding /heterozygosity deficit/ coefficient of an individual within the total population (FIT) in the locus are 0.03, 0.08 and 0.11, respectively. Pairwise population Fst are also indicated in Table S 42 with the highest value (0.12) for Hugub to Banja and Shubi Gemo to Hugub populations. The highest pairwise Fst (0.12) is also found between SA2 and A1 agroecologies (Table S 43). Structure analysis (K = 2) supports two different gene pools (Figure 12; Figure 13).From the entire populations, only Kumato populations meet the assumption of Hardy Weinberg Equilibrium (HWE). Unrooted neighbor-joining tree topology indicated a clear IC subgrouping into two distinct clusters (Figure 8). Even though the reported rate of gene flow (Nm) was minimum (8.6), high rate of admixture was noted among the predetermined populations. The mantel correlation test across population genetic and geographic distance shows no correlation (Figure 11).     Among Pops 5%Among Indiv 12%Within Indiv 83%Among Pops 3% Among Indiv 8%Within Indiv 89%     The 23 to 30 position downstream of the repeat region was sequenced as \"ATTTTGAG\", whilst, 3 alleles sequences were found to have different repeats than respective reference sequences. 24 and 1 SNP substitutions were found at positions 39 and 46, respectively. 12 insertion SNPs and 2 deletions were noted on the upstream polymorphism positions of -30 to -29 positions. Besides, 2 nucleotide substitutions were reported at -61 upstream polymorphism, while, 3 substitution at -28 position. The consensus sequence size deviation from fragment size ranged from 1 to 115 bp.Further polymorphisms were also observed in different positions of the repeat structure other than the positions considered hereunder. B10, B11.1, B13, B72 haplotypes were obtained from the allelic sequence. The invariable (monomorphic sites) and variable (polymorphic) sites found were 412 and 35, respectively. A total of 26 singleton sites and 5 parsimony informative sites were observed from the package of DNA sequence polymorphism while the total number of mutation sites and indel events were 33 and 17, respectively. The number of indel haplotypes and indel diversity were 5 and 0.00465). Haplotype diversity (Hd) was 0.82. The haplotype-based phylogenic analysis using Neighbor-Joining (NJ) showed that indigenous chicken populations are mainly clustered into two gene pools comprising different subpopulations as obtained from the structure analysis of allele sizes from capillary electrophoresis. \uD835\uDC4E The codes of chicken populations; \uD835\uDC4F Δ Defined deletion compared with the reference sequence. ∘ is consistent with the reference sequence; * unique alleles  Neighbor-Join and BioNJ algorithms to a matrix of pairwise distances estimated using the Maximum Composite Likelihood (MCL) approach and then selecting the topology with superior log-likelihood value. A discrete Gamma distribution was used to model evolutionary rate differences among sites (2 categories (+G, parameter = 0.0500)). The tree is drawn to scale, with branch lengths measured in the number of substitutions per site. The analysis involved 10 nucleotide sequences. Codon positions included were 1st+2nd+3rd+Noncoding. There were a total of 444 positions in the final dataset.In this study, we report polymorphism at the MHC-B microsatellite marker, LEI0258, in indigenous chicken village chicken populations from Ethiopia. From the 29 allele size reported here, 22 are novel alleles (Fulton et al., 2006;J. E. Fulton et al., 2016;Gupta et al., n.d.;Izadi et al., 2011;Keambou et al., 2014;Lwelamira et al., 2008;Ncube et al., 2014;Nikbakht et al., 2013).The number of alleles reported was higher than the numbers reported by different authors in previous works on indigenous chicken populations (Han et al., 2013;Nikbakht et al., 2013) The high diversity at the marker may be a direct consequence of the diversity of disease challenges facing Ethiopian chicken within and across different agro-ecologies, with polymorphism at the locus maintained by balancing selection with high LEI0258 diversity increasing the diversity of antigens being presented to T-cells (Chazara et al., 2013).Overall, 50% of the alleles in this study are only found in only two or three of the populations out of the 24 considered. With the exception of alleles, 197, 312, 315 and 351 bp, the remaining of the alleles occurred at a lower frequency (< 0.20). The low-frequency abundance of LEI0258 alleles might be attributed to a new mutation arising in a population and therefore available only in a few individuals. It might also be due to its susceptibility to disease and other selection pressures resulting unfit to survive the production challenge the chicken is facing in the variable environments where they were sampled. High frequency of allele 315 at Batambe population could imply a fitness or survival advantage to the individual carrying it resulting in it being selected for and occurring at higher frequencies (31%). This, however, needs to be further study in absence of any information regarding the possible association between disease resistance/susceptibility and the allele.We did not identify here those alleles which previously have been shown to be a positive correlation with NDV (206 bp) and body weight (307 bp) ((Fulton et al., 2006;Fulton et al., 2016 andLwelamira et al. 2008) For the later it may not be surprising considering the small size of Ethiopian village chicken compared to their commercial counterparts; while the former is suggest that allele (206 bp) may not be of relevance to NDV resistance/susceptibility in Ethiopian village chicken. Several LEI0258 alleles were shared among the predetermined populations implying that they have been subjected to either directional selection or due to recombination effect (Miller et al., 2004;Nikbakht et al., 2013;Salomonsen et al., 2014). Alleles 206 and 307 bp, reported in Tanzanian chicken to associate with Newcastle disease antibody response and body weight, respectively, were not reported in Ethiopian chicken in this study (Lwelamira et al., 2008). The highest mean genetic distance (d = 0.94) was between Alfa Midir and Batambe followed by the former and Gafera (d = 0.90) populations. Overall, the genetic distances found in this study are small indicating that these populations are more related and have a common ancestor. Considering this result it looks that the genetic diversity pattern fragments these populations west to east from North to South geography line. This could be attributed to their proxy of entry history of chicken introduction to Ethiopia/ Africa. The average Fst for the entire population was 0.028 ascertaining that 2.8% of the total genetic variation corresponded to the differences among populations, whilst 97.2% was explained by differences among individuals. The smallest genetic differentiation between some populations is justified by minimal selective breeding, uncontrolled mating and movement of a live chicken.A subset of allele sequencing result ascertained the presence of single nucleotide polymorphisms (SNPs) in LEI0258. The two main VNTR were the R13 and R12. R13 with a 13 bp repeat unit, \"CTATGTCTTCTTT' was found with a frequency of only once similar to Wang et al. (2014) and inconsistent with other studies with more frequencies (Chazara et al., 2013;Nikbakht et al., 2013).The 23 to 30 position downstream of the repeat region was sequenced as \"ATTTTGAG\". They agree with the sequences obtained by Fulton et al. (2016), Han et al. (2013), andIzad et al. (2011).Twenty-four and one SNP substitutions were found at positions 39 and 46, respectively. The number of R12 motifs (CTTTCCTTCTTT) in the individual sequences ranged from 2 to 18, whilst, only one R13 motif was found. 12 insertion SNPs and 2 deletions were noted on the upstream polymorphism positions of -30 to -29 positions. Besides, 2 nucleotide substitution were reported at -61 upstream polymorphism, while, 3 substitution at -28 position. 6 allele sequences were found to have different repeats than respective reference sequences (ATTTTGAG). Results of allele size from both fragment length and consensus sequences did not exactly much which might be because of the difference in environmental factors, technological approaches, and precisions. The size deviation ranged from 1 to 115 unlike what was reported by Han et al. (2013) who found size differences of 1 to 69 range. The comparison between the fragment sizes and consensus sizes (bp) across population consistently showed higher values for the later except for the Dara population.This result was not consistent with the works of Fulton (2016) and Han et al., (2013). Further polymorphisms were also observed in different positions of the repeat structure other than the positions considered hereunder. Only a few haplotypes (B10, B11.1, B13, B72), were found when compared with the haplotypes reported by Fulton et al. (2016) and Chazara et al (2013). In contrast, none of these haplotypes were reported in the report of Wang et al. (2014) for Chines chicken and Ngeno et al. (2015) for Kenyan indigenous chicken.The SNP based phylogenic analysis using Neighbor-Joining (NJ) showed that indigenous chicken populations are mainly clustered into two gene pools comprising different subpopulations as obtained from the structure analysis of allele sizes from capillary electrophoresis. From phylogenetic analysis, clear separation of ecotypes was not noted indicating genetic admixture between populations.Very high diversity was found in Ethiopian indigenous chicken populations at LEI0258, this diversity is observed within all population. Our results support the importance of MHC diversity in response to the disease challenges faced by smallholder poultry production in Ethiopia.Breeding improvement programs will need to maximize this diversity through balancing selection that maintains polymorphisms and increases within-population diversity. This very high diversity report for Ethiopian indigenous chicken populations on LEI0258 locus will provide a framework for the existing and future chicken breed improvement interventions. Besides, we can infer that the genotyping of the tandem repeat microsatellite marker LEI0258 is a suitable method for MHC typing of indigenous Ethiopian chicken populations considering the high level of polymorphism observed at the locus within and across indigenous populations. Polymorphisms from the sequencing result, also support the genome diversity of indigenous Ethiopian village chicken populations. As a way forward, studying the relationship of these polymorphisms and the disease resistance/susceptibility haplotypes in Ethiopian chicken populations should be undertaken.Indigenous chicken makes a profound contribution to the rural economies in Ethiopia by playing a major role for the rural poor and marginalized people as a subsidiary income and consumption.Genetic diversity represents the total genetic variation among populations and several measures of diversity have been developed over years (Barrandeguy and García, 2014). Sufficient genetic variation in livestock populations is necessary both for adaptation to future changes in climate and consumer demand and for continual genetic improvement of economically important traits (Aslam et al., 2012;Eggen, 2012;Schmid et al., 2015). To understand phenotypic variation in farm animals such as in poultry, it is essential to define all potential genomic variation within a genome (Schmid et al., 2015). Evolution of chickens and programs for their artificial selection rely on the availability of sufficient levels of genetic variation. In response to the global shift in environmental conditions and market demands for chicken products, the diversity of village chicken is needed for future improvements programs (Muchadeyi et al., 2008).Different techniques are involved to study genetic variations., DNA variation known as SNPs, the most abundant sources of genome variation, have become increasingly markers of choice in genomics studies (Alderborn, 2000;Gheyas et al., 2015) with genome annotation, the link between biological or functional information and genome sequences an important gain in our understanding of how genomic variations influence phenotypes (Fulton, 2012). Previous studies on Ethiopian chicken failed to comprehensively characterize the chicken genomic diversity in Ethiopia. Only a few Ethiopian indigenous chicken populations have been characterized using molecular markers. These are Tillili (Alemayhu, 2003;Mogesse, 2007), Jarso, Horo, Chefie, Tepi (Tadelle et al., 2003), Gellila, Debre Ellias, Melo Hamusit, Farta (Mogesse, 2007), Sheka, Konso, and Mandura ( (Desta et al., 2014) chicken populations. With regard to the mapping of genes of interest in indigenous Ethiopian chickens, so far a single work has been published (Wragg et al., 2012). Wragg et al. (2012) studied randomly selected 15 birds from 5 Ethiopian chicken populations (in Gondar, Konso, Gumuz, sheka, and Guduro). Together with other breeds, the study mapped phenotypic traits (e.g. skin and egg color) using genome-wide association approaches. It illustrates for the first time the possibility and power to use indigenous outbreed chicken population for the fine genome mapping of Mendelian traits. Wragg and his colleagues also recommend including, a minimum of 90-110 kb SNPs for effective genome-wide associations study in village chickens.The presence of genomic diversity in domestic chicken is of great importance and a prerequisite for rapid and accurate genetic improvement of selected breeds in various environments, as well asto facilitate rapid adaptation to potential changes in breeding goals (Nielsen, 2005). Hanotte et al.(2010), suggested that it is time to tap Africa's livestock genomes to better understand and exploit the genetic diversity of Africa's individual livestock breeds before they fade away. Understanding which factors shape levels of genetic diversity within genomes forms a central question in evolutionary genomics and is of importance for the possibility to infer episodes of adaptive evolution from signs of reduced diversity. There is an on-going debate on the relative role of mutation and selection in governing such diversity levels (Mugal et al., 2013). This study aims at characterizing genomic diversity through the discovery of genomic variants in 27 indigenous chicken populations of Ethiopia as a milestone for further works on management and conservation of chicken genetic resources, studying the genetic mechanism underlying local chicken adaptation.Blood samples were collected from 27 chicken populations in Ethiopia (Figure 16). Samples included 103 cocks and 157 hens. Except the improved Horro, Meseret, Tsion Teguaz, Jarso and local Horro populations, 10 chicken from each village were sampled. One or two chicken were sampled per household. Improved Horro was sampled from a breeding stock of 8 th generation under selection at Debre Zeit Agricultural Research Centre and used as a reference population.Unlike other populations, Jarso and Local Horro sequences were obtained and included from the previous studies. Photographs and weight of each bird were taken. The average weight of sampled chicken was 1.26 kg with age ranges of 5 to 36 months. Sampling considered different agroecological zones, altitudes ranging from 729-3500 meters, marketing points, and chicken phenotypic characteristics. From the wing vein of each chicken, 50 -250 µl of whole blood were drawn with syringes using cryotubes filled with 1.5 ml absolute ethanol (100%) following the guidelines available at https://www.sheffield.ac.uk/nbaf-s/protocols_list. Total DNA was extracted from chicken whole blood at the BecA-ILRI Hub, Nairobi, Kenya facility (http://hub.africabiosciences.org/) using the Qiagen DNeasy blood and tissue kit protocol (Lwelamira et al., 2008). To evaluate the DNA concentration a Thermo Scientific NanoDrop spectrophotometer 2000c was used. The integrity of DNA was confirmed by agarose gel electrophoresis whereby 20 ng/µl genomic DNA samples were loaded with 1 µl loading dye (6X) on a 1% agarose gel containing 2.5 µl gel red at a voltage of 7/cm for 60 minutes, 3 µl of lambda DNA of size of 48,500 bp and a concentration of 20 ng/µl was used as size marker and the gel was then examined using UV light using GelDoc-It2 Imager to check the extracted DNA quality and quantity. The genomic DNA from (n = 284) was normalized to a final volume of 100 µl and final concentration of 50 ng/µl and sent to Edinburgh Genomics, UK, for whole genome sequencing.The QC, library prep, and sequencing were performed at the Edinburgh Genomics facility.Genomic DNA (gDNA) samples were evaluated for quantity and quality using an AATI (Agilent The libraries were evaluated for mean peak size and quantity using the Caliper GX Touch with a HT DNA 1k/12K/HI SENS LabChip and HT DNA HI SENS Reagent Kit. Those libraries were then normalized to 5 nM using the GX data and the actual concentration was established using aRoche LightCycler 480 and a Kapa Illumina Library Quantification kit and Standards. The normalized libraries were denatured and pooled in eights for clustering and sequencing using aHamilton MicroLab STAR with Genologics Clarity LIMS X Edition. Libraries were clustered onto HiSeqX Flow cell v2.5 on cBot2s and the clustered flow cell is transferred to a HiSeqX for sequencing using a HiSeqX Ten Reagent kit v2.5. The samples were sequenced at a genome coverage of ~5-90X (mean = 36.1X). Demultiplexing is performed using bcl2fastq (2.17.1.14), allowing 1 mismatch when assigning reads to barcodes.Adapters (Read1: AGATCGGAAGAGCACACGTCTGAACTCCAGTCA, Read2:AGATCGGAAGAGCGTCGTGTAGGGAAAGAGTGT) are trimmed during the demultiplexing process.The pipelines for mapping and variant calling included: mapping reads against reference genome using BWA-mem, sorting BAM file, removing duplicated reads with PICARD, Base Quality Score Recalibration (BQSR) with GATK, calling variants using GATK, Variant Quality Score Recalibration (VQSR) for variant filtration with GATK, and finally selection of only bi-allelic SNPs which passed the VQSR step.High quality paired-end reads (FASTAQ format) were aligned to the chicken (Gallus gallus) reference genome sequence (Gallus_gallus-5.0 or galGal5)(https://www.ncbi.nlm.nih.gov/genome/?term=Galus+galus+5), using Burrows-Wheeler Aligner software package (http://sourceforge.net/projects/bio-bwa/files/) with the command 'mem -t 8 -k 32 -M -R' ( where -t = no. of threads; -k = min seed length; -M = Mark shorter split hits as secondary (for Picard compatibility) which permits high-quality queries for longer sequences as it is fast and accurate (Li and Durbin, 2010); and -R for defining read groups. The alignment output generated were stored in the SAM format and then converted to BAM formats using PICARD tools. Duplicated reads originating from a single fragment of DNA during sample preparation (such as library construction using PCR) were marked and removed using PICARD's MarkDuplicates command (https://broadinstitute.github.io/picard/command-lineoverview.html#MarkDuplicates).BQSR is a data pre-processing step that detects systematic errors made by sequencers in estimating the quality score of each base call. Base quality score is an important parameter for variant calling as it expresses confidence that the base has been called correctly. Unfortunately, the scores produced by the machines are subject to various sources of systematic technical errors, leading to over-or under-estimated scores. The BQSR step applies a machine learning algorithm to model these errors empirically and adjust the quality scores accordingly by considering a number of covariates such as sequencing context of the base, position in read or sequencing cycle (https://gatkforums.broadinstitute.org/gatk/discussion/44/base-quality-score-recalibration-bqsr).Variant calling from each sample was performed in the gVCF mode for cohort analysis using GATK's HaplotypeCaller (Figure 2). Joint genotyping of samples from each population were done using GATK's GenotypeGVCF tool for downstream analysis. Variant Quality Score Recalibration (VQSR) were also performed to increase sensitivity (identifying the real variants) and specificity (identifying false positives) using GATK followed by a selection of only bi-allelic SNPs that passed the VQSR step. For the VQSR step, we used 1M validated SNPs and 15 SNPs from dbSNP for recalibration purpose.Figure 17. Overview of data analysis pipeline using BWA, PICARD, and GATK.Population structure and relationships between samples were established using Principal Component Analysis (PCA) using smartpca program in eigenstrat version 6.0 (Patterson et al., 2006;Price et al., 2006). PCA was performed using all SNPs from 27 populations (n = 20,867,451 SNPs). The top three principal components (PCs) provided the clearest separation of the data and were used to construct the PCA plot. Apart from the PCA, the genetic structure of each population was also assessed unsupervised, using ADMIXTURE version 1.3.0 (Alexander et al., 2009).Global admixture analyses were run for K = 2 to K = 10 assumed numbers of ancestors using 651,417 LD pruned SNPs. The optimal K value was determined based on the lowest crossvalidation error. The following criteria was used for LD-pruning: \"indep-pairwise 50 10 0.3\", \"bpspace 1000\", \"maf 0.1\" and \"geno 0.1\". This command targets for removal each SNP that has an r 2 value of greater than 0.1 with any other SNP within a 50-SNP sliding window (advanced by 10 SNPs each time) and Minor allelic frequencies of 0.1. The average genome nucleotide diversity (π) for each population was determined using VCFtools version 0.1.13 in 20 kb windows over a 10 kb sliding step (Danecek et al., 2011). Structure and admixture plots where plotted using Rsoftware package version 3.4.3.To predict their functional consequence, SNPs were annotated against the Ensembl chicken gene database (release 92) using the software package ANNOVAR (Wang et al., 2010). The effects of non-synonymous SNPs on protein function were predicted based on evolutionary conservation using the Sorting Intolerant from Tolerant (SIFT) prediction algorithm which depends on the degree of conservation at individual amino acid (AA) positions (Sim et al., 2012). Using multiple alignments of homologous but distantly related peptide sequences, SIFT calculates normalized probabilities (SIFT score) of observing all possible AA residues at a position (Gheyas et al., 2015;Ng, 2003). If the SIFT score is greater or equal to 0.05 the variant is considered evolutionary tolerant (TOL), whereas variants with a score less than 0.05 are regarded as intolerant (INTOL) and potentially deleterious (Choi and Chan, 2015;Kumar et al., 2009;Sim et al., 2012). The SNPs were further checked for their overlap with 1.1 million conserved elements (CE) obtained from the Roslin Institute (Eory et al., unpublished data; personal communication). These CEs were generated from multiple alignments of sequence data from 48 bird species and 1 anole lizard species using Genomic Evolutionary Rate Profiling (gerp++) (Davydov et al., 2010).The proportion of homozygous SNPs were calculated using the \"stat\" option of the VCFtools version 0.1.113 (Danecek et al., 2011). To establish the biological significance of a list of genes carrying potential functional SNPs, the DAVID Bioinformatics Resources 6.8 (DAVID; Huang et al. 2009aHuang et al. , 2009b) ) was used. This allowed performing enrichment analysis of the Gene Ontology (GO) and the Kyoto Encyclopaedia of Genes and Genomes (KEGG) pathways (KOBAS version 3.0, http://kobas.cbi.pku.edu.cn/). For both analyses, the Fisher exact P-value < 0.05 default threshold were considered.The average number of paired sequence reads generated from an individual sample in each population ranged between to about 202 million (Hadush Adi) and 475 million (Surta), resulting in average genome coverage of about 22X to 44X. On average 88% of the bases were covered by at least 5 reads and > 90% were covered with at least 10 reads (Table 11). More than 98% of read pairs in all samples were mapped to the Galgal 5.0 reference genome. The mean sequence depth of the entire chicken population sampled is about 39X (Figure 18).Variant calling and filtration resulted in the detection of about 21 million SNPs (n = 20,867,451) combining the 27 population. The number of SNPs detected within individual population ranges from 10 to 12 million. The mean SNP density reported is 21 SNPs ± 5 per kb or 1 SNP for every 48 bases. 28.12 % (n = 5,868,599) of the SNPs were already reported in dbSNP (build 147), which currently contains ~21 million SNPs (n = 20,867,451 SNPs) for chicken; while the rest of the SNPs (n = 14,998,852) are novel. Much lower SNP density is found in sex chromosomes than autosomes.The genome landscape plot for the genes for individual chromosome (1 kb window) is given in Widespread variations are noted in SNP density across chromosomes (Figure 20). The lowest SNP density was reported for chromosome W and the highest SNP density was reported for chromosome 26 followed by chromosome 6. Chromosome-wise SNP distribution plots across the Ethiopian indigenous chicken genome are depicted in Figure 19. The peak number of non-synonymous SNPs were observed in chromosome 16. Except for Jarso, about 11% to 15% of the SNP are novel in each chicken population (Table 12). The private allele peak is owned by the Jarso chicken (n = 280448) and the lowest by the Batambe chicken (n = 34532). Alternate allele frequency ranges from 0.35 for Loya to 0.39 for Local Horro and Hugub chicken populations.About 18 -19% of the SNPs overlap conserved elements. The population with a relatively highest putatively functional variant is Kumato followed by Local Horro chicken. The proportion of heterozygote SNPs is 55.21 % for the overall chicken population (n = 284). Bekele Girrisa chicken population has the highest average heterozygous SNPs (60%) followed by Kumato chicken (59%).The lowest average heterozygous SNPs is reported in Hugub chicken population (about 40%). The mean genome nucleotide diversity (π) for the entire chicken population is 0.02 ± 0.001. The average transition to transversion ratio is 2.35. The highest number of nucleotide substitutions was recorded for C > T and G > A (4 X 10 6 ; Figure 21).     Annotation of 21 million SNPs against ENSEMBL gene annotation database shows that 46.36% of SNPs are located within genes (intronic + exonic + UTRs + splicing) and the rest are found outside genes (intergenic and up/downstream). However, only 1.6% (n = 331,968) of the SNPs are in protein-coding regions (i.e. exon) (Table 13). SNPs in exonic regions were further classified into synonymous (65.14%; variants which do not alter the amino-acid sequence in proteins), non-synonymous (34.38%; variants that change amino-acid sequence in a protein) and stop gain or loss (0.98%; variants that leads to gain or loss of stop codon) polymorphisms (Figure 26). The synonymous and non-synonymous (AA-altering) number of SNPs are 190,041(0.48%) and 100,293 (0.98%), respectively. Whereas, the other AA altering variant, number of stop gain/loss accounts for about 0.36% (n = 1,209). Even though non-synonymous SNPs change amino acid sequence within a protein, the effects are not always harmful or radical on protein function. Using SIFT, 21.9% of the non-synonymous variants (n = 44, 553) were predicted as 'intolerant' (INTOL) having a radical effect, 64.94% (135,917) were predicted 'tolerant' (TOL), whereas the prediction for other variants had low confidence level (Figure 27).The average percentage of homozygous and heterozygous SNPs (%) in Ethiopian indigenous chicken populations.Table 12 shows the number and percentage of putatively functional SNPs that are present in high frequency and were detected from 27 populations.Apart from the amino-acid altering variants, other potentially functional categories reported in this study are splicing variants (0.006%), variants in 3′ and 5′ UTR regions with possible roles of regulating protein translation (0.82%); those within 1 kb up-or downstream of transcription start or end sites (3.06%) with possible roles on transcriptional regulation; and finally, the SNPs belonging to the non-coding RNA (ncRNAs) (2.61%) (Table 13). Nonsynonymous number of SNPs/Kb across chromosomes are presented in Figure 25.  Upstream: a variant that is located in the 1-kb region upstream of the gene start site; stop gain: a non-synonymous (ns) SNP that leads to the creation of a stop codon at the variant site; stop loss: a non-synonymous SNP that leads to the elimination of a stop codon at the variant site; splicing: a variant within 2 bp of a splice junction; downstream: a variant that is located in the 1-kb region downstream of the gene end site.  The frequency spectrum of non-reference or alternative alleles (AAF) of variants from different annotation categories were compared (Figure 28; Figure 29). The allele frequency distribution of different annotation categories showed that the largest proportion of variants fell within the AAF frequency bin of ≤ 10%. However, the proportion of low-frequency variants was higher for potentially harmful variants like deleterious missense and stop gain/loss (> 60%) compared to variant belonging to the neutral categories (e.g. intergenic, intronic, and synonymous; < 40%).Polymorphisms that are potentially functional or deleterious but are present at high frequency (e.g. AAF > 0.9) may be under selection. Table 4 shows the number and percentage of putatively functional SNPs present in high frequency and were detected from 27 populations. One thousand four hundred twenty-five functionally annotated SNPs identified in the entire dataset (27 populations) were checked for their functions and biological pathways (Table S 45). GO term enrichment analysis shows significant (P ≤ 0.05). GO terms related to innate antibacterial and antifungal immunity response (IPR000157; Toll/interleukin-1 receptor homology (TIR) domain, IPR007110); Immunoglobulin-like fold and energy biosynthetic ((GO: 0016887; ATPase) activity, (GO: 0006183; GTP biosynthetic process), (GO: 0006228; UTP biosynthetic process), GO: 0006241; CTP biosynthetic process)). Similarly, 385 non-synonymous deleterious genes (352 reported) detected in 10 populations (Table S 46) where functionally annotated and GO term enrichment analysis gave genes responsible mainly for DNA repair and binding (GO:0042162, GO:0006281 ), ATP binding (GO:0005524) and WD40 repeat domains (IPR017986, IPR015943, IPR001680) (Table 15). Annotation of non-synonymous deleterious SNPs with AAF > 0.9 from 27 populations (Table S 47) and their functional characterization has confirmed genes attributed to methyltransferase, protein autophosphorylation, a class of nuclear body called promyelocytic leukemia (PML) (Table 16).   Single nucleotide polymorphisms (SNPs) and other mutations may disrupt the RNA structure by interfering with the molecular function and hence cause a phenotypic effect (Sabarinathan et al., 2013). In this study, we performed whole-genome sequencing for SNPs and used the identified SNPs to characterize genetic diversity in indigenous chicken populations of Ethiopia. About 21 million (n = 20,867,451) high quality SNPs were discovered in 27 populations (n = 284 birds).The 21 million SNPs discovered in this study were higher than the number of SNPs discovered in a previous study by Gheyas (2015)  this study reported comparatively in concordance with the previous study by Gheya et al., (2015) who reported 16 ± 10.01 SNPs/Kb. In contrast to the findings of Gheyas et al. (2015), macro chromosomes where found to have a higher number of SNP density despite high recombination rate and in turn high SNP polymorphism in micro chromosomes (Burt, 2005). Even though the high gene-density of the smaller chromosomes would make them susceptible to hitchhiking effects that could erode genetic variation, these effects appear to be offset by the far higher recombination rate of the microchromosomes (Aslam et al., 2012).Much lower SNP density is found in sex chromosomes than autosomes as the high number of repetitive sequences mainly in W chromosome, whereas, the highest SNP density was reported for Chromosome 26 followed by chromosome 6 in spite of their lowest chromosomes size unlike what is reported by Gheyas et al. (2015). The high density of SNPs were obtained for chromosome Z in the current reference genome (11 SNPs/kb) compared to an average of ~3 SNPs/kb reported by Kranis et al. (2013). This specific finding is in line with the statement of Wong et al. (2004) who stated that SNP density is independent of chromosome size. The second lowest SNPs/Kb detected from Chr31 could be mainly because of the partial representation of this chromosome in the current reference genome (Kranis et al., 2013). Our result, also is not in line with the findings of previous studies which reported reduced genetic variations (much lower number of SNPs) on ChrZ than on autosomes for a multitude of potential reasons like low male effective population size due to skewed reproductive success among males leading to male effective population size, selective sweep due to selection on sex-linked characters combined with lower recombination rates (Sundstrom, 2004). Despite, lower chromosome length, chromosome 16 was also found to have a higher number of SNPs (21 ± 5) though this chromosome is an exception to the characteristic of high recombination polymorphism rates in micro-chromosomes. This may be due to the presence of highly variable MHC genes Whereas, the W chromosome is the sex-limiting chromosome in chicken and its reduced genetic variability is the result of selection and the complete lack of recombination outside the pseudo autosomal region (Berlin and Ellegren, 2004;Moghadam et al., 2012;Wright et al., 2016;Xu et al., 2018).The peak number of non-synonymous SNPs were observed in chromosome 16. Even though nonsynonymous SNPs change amino acid sequence within a protein, the effects are not always harmful or radical on protein function. Much higher SNP density in Chr16 (probably because it contains highly variable MHC regions) and also in smaller chromosomes (chr25-33). A smaller chromosome may be gene rich and hence you see greater SNP density. About 18 to 19% of the SNPs overlapped the conserved elements. The population with a relatively highest putatively functional variant is Kumato followed by local Horro chicken population. The proportion of SNPs that were found heterozygous 55.21% for the overall chicken population (n = 284) is higher than the average heterozygous SNPs recommended (50%). The mean genome nucleotide diversity (π) of the entire chicken population is 0.02 ± 0.001 reported in this study lies in the range reported by Lawal (2018). The lowest average heterozygous SNPs is reported for Hugub chicken populations (about 40%). The local habitat of this specific population is Hot to warm semi-arid lowlands. The lowest heterozygous SNPs could be attributed to the fact that as the area is extremely hot and low density of the chicken population (narrow breeding base). The average transition to transversion ratio of detected SNPs is 2.35 (Aslam et al., 2012). The expected Ti/Tv ratio of true novel variants can vary across the genome attributed to a variability in the CpG and GC content of the genome.For instance, in the case of exomes, an increased presence of methylated cytosine in CpG dinucleotides in exonic regions leads to an increased Ti/Tv ratio due to an easy deamination and transition of a methylated cytosine to a thymine. It is also observed that GC content is higher in birds and mammals than in invertebrates. Observed Ti/Tv ratio in this study is lower than the findings from Alsam et al. (2012) (2.45). This finding is in contrast to the fact that birds have higher TS/TV ratio for owning smaller genome size and a higher GC percentage in bird genomes.Principal component and admixture analyses suggest the presence of four ancestral gene pools across the populations. The close proximity of majority of the populations often reflected their geographic proximity. The clustering of these population follows the geographical pattern where they are sampled from.The functional information of these variants can help in the prediction of phenotypes or genetic merit with higher accuracy and selection of individuals can be done accordingly. Annotation of 21M SNPs against ENSEMBL gene annotation database shows that 46.36% of SNPs are located within genes (intronic + exonic + UTRs + splicing) and the rest are available outside genes (intergenic and up/downstream), while, only 1.6% (n=331,968) of the SNPs are in protein-coding regions (i.e. exonic). However, the study by Wong et al. (2014) showed that only ~37% of the variants fell within genes with only 1.2% fell within the coding regions. Non-coding RNAs are an important class of genes, responsible for the regulation of many key cellular functions (Cao, 2014;Frías-Lasserre and Villagra, 2017;Gardner et al., 2015). The highest number and percentage of putatively functional SNPs that are present in high frequency and were detected from 27 populations shows that these genes are adaptive.SNPs in a coding region can be synonymous (do not result in a change in amino acid; selectively neutral) and non-synonymous. The synonymous and non-synonymous (AA-altering) number of SNPs are 190,041(0.48%) and 100293 (0.98%), respectively. Whereas, the other AA altering variant, number of stop gain/loss accounts for about 0.36% (n=1,209). Even though nonsynonymous SNPs change amino acid sequence within a protein, the effects are not always harmful or radical on protein function. Using SIFT, 21.9% of the non-synonymous variants (n = 44, 553)were predicted as 'intolerant' (INTOL) having a radical effect, 64.94% (135,917) were predicted 'tolerant' (TOL), whereas the prediction for other variants had low confidence level (Figure 11).Much higher SNP density in Chr16 (probably because it contains highly variable MHC regions)and also in smaller chromosomes (chr25-33). A smaller chromosome may be gene rich and hence may have greater SNP density. Apart from the amino-acid altering variants, other potentially functional categories are also reported in this study, such as splicing variants (0.006%); variants in 3′and 5′ UTR with possible roles of regulating protein translation (0.82%); those within 1 kb up-or downstream of transcription start or end sites (3.06%) with possible roles on transcriptional regulation; and finally, the SNPs belonging to ncRNAs (2.61%).Genomic regions conserved across distantly related species are assumed to be under purifying selection, and hence variants within these regions are likely to be harmful (Gheyas et al., 2015).Hence, SNPs overlapping evolutionarily conserved elements were checked as these may have potential functional effects. The 21 million SNPs were annotated against 1.1 million conserved elements (CEs) across 48 birds plus a lizard. These CEs covers about 2.1% of the chicken genome (total length of CEs is 186,488,363 bases).The allele frequency distribution of different annotation categories showed that the largest proportion of variants fell within the AAF bin of ≤ 10%. However, the proportion was higher for potentially harmful variants like deleterious missense and stopgain/loss (> 60%) compared to neutral categories like intergenic, intronic, and synonymous (< 40%). This is expected as potentially detrimental SNPs are expected to be mostly low frequency. However, contrary to our expectation, we did not find any variation in the AAF pattern of SNPs within the CE category with potentially neutral variants. SNPs that are potentially function or deleterious but are present in high frequency (e.g. AAF > 0.9) is expected to have greater impact and may be under selection.Seven hundred ninety-five reported functionally annotated genes found from the entire 27 population were extracted and checked for their functions and biological pathways with the highest stringency. GO terms related to innate antibacterial and antifungal immunity response (IPR000157; Toll/interleukin-1 receptor homology (TIR) domain, IPR007110); Immunoglobulinlike fold and energy biosynthetic ((GO: 0016887; ATPase) activity, (GO: 0006183; GTP biosynthetic process), (GO: 0006228; UTP biosynthetic process), GO: 0006241; CTP biosynthetic process)). Toll proteins or Toll-like receptors (TLRs) and the Interleukin-1 receptor (IL-1R)superfamily are both involved in innate antibacterial antifungal, anti-Protozoan and anti-viral immunity in chicken, insects and in mammals (Blasius and Beutler, 2010;Cohen, 2014;Liao et al., 2010;Liu and Zhao, 2007;Ma et al., 2007). Interleukin-1 receptor family participates in the regulation of immune responses, inflammatory reactions, and hematopoiesis (Armant, 2002;Beutler, 2004;Mukherjee et al., 2016;Takeda and Akira, 2001;Vasselon, 2002). Protein protease inhibitors constitute a very important mechanism for regulating proteolytic activity. Annotation of non-synonymous deleterious SNPs with AAF > 0.9 and their functional characterization has confirmed genes attributed to Methyltransferase, Protein auto-phosphorylation, a class of nuclear body called promyelocytic leukemia (PML) which react against SP100 auto-antibodies during viral infections; and a cell aging process associated with the dismantling of a cell as a response to telomere shortening and/or cellular aging and genes that controls positive regulation of DNA damage response, signal transduction by p53 class mediator. The VPS36 gene which is also known to have a plausible function in comb characters have been discovered as non-synonymous variant in the 27 indigenous chicken populations of Ethiopia (Shen et al., 2016b). Similarly, the MX1 gene responsible for morbidity, early mortality, viral shedding, and cytokine responses have been foundin nonsynonymous variants of the 27 population. It is an interferon-induced gene which inhibits the proliferation of single-stranded RNA viruses (Fulton et al., 2014;Mishra et al., 2011;Schusser et al., 2011;Selvaramesh et al., 2018;Wang et al., 2012). The gene Chicken Aggrecan Core Protein (ACAN) which has an association with Tibia Dyschondroplasia is also found harboring non-synonymous deleterious variants in 10 populations (Fan et al., 2013;Stattin, 2009).The genetic make-up of populations is the result of a long-term process of adaptation to specific environments, ecosystems and of artificial selection. Selective breeding for genetic improvement is expected to leave distinctive selection signatures within genomes. The identification of a signature of selection can help to elucidate the mechanisms of selection and accelerate genetic improvement by well understanding molecular pathways underlying phenotypic traits and breeding goals (Elferink et al., 2012). Selection leads to specific changes in the patterns of variation among selected loci and in neutral loci linked to them (Guo et al., 2016). These genomic footprints of selection are termed as signatures of selection and usually used as to identify loci that have been subjected to selection. Various statistical approaches, either the allelic frequency spectrum or the properties of haplotype segregation in populations are being used for detecting selection signatures at genome-wide scale (Qanbari et al., 2015). Among others, pooled heterozygosity (Hp) statistic is a variability indicator based on allele counts across sliding windows. The other commonly used statistic is the fixation index (Fst), which measures the genetic differentiation based on variations in allelic frequencies among populations (Qanbari and Simianer, 2014). The loci in the tails of the empirical distribution of Fst are the candidate targets of selection (Akey, 2002). The evolution of new functions and adaptation to new environments occurs by positive selection, whereby beneficial mutations increase in frequency and eventually become fixed in a population (Tang et al., 2007). Local environmental adaptation and artificial selection can change the allele frequencies at specific loci, leading to a higher level of population differentiation (Fst) (Yang et al., 2014). Adaptation or positive natural selection leaves an imprint on the pattern of genetic variation found in a population near the site of selection (Xue et al., 2009).Local breeds make up most of the world's poultry genetic diversity and are still very important in developing countries where they represent up to 95 percent of the total poultry population. These local breeds, which are well-adapted to extensive husbandry systems and suitable for resourcepoor poultry farmers endowed with very limited means, should be thoroughly studied as a basis for enhancing their use and conservation (Besbes et al., n.d.). To understand phenotypic variation in farm animals and in poultry, in particular, it is essential to define all potential genomic variation within a genome (Schmid et al., 2015). Discovery of genes with large effects on economically important traits has for many years been of interest to breeders (Wolc et al., 2014). In this study, we used Hp and Fst statistics to detect the signature of selection in improved and other indigenous chicken populations.Recent advances in sequencing technologies have helped in the detection of candidate genome regions playing crucial roles in the evolution of production and reproduction traits in chicken. In this regard, various genes responsible for growth and egg production has been found. This study aims to elucidate the effect of on-station improvement on the signature of positive selection in an indigenous chicken, the Improved Horro, compared to non-improved indigenous chicken (Local Horro, Hugub, Arabo, and Jarso).123A well as for the later on their cumulative egg production 24 weeks after the start of laying (EN24).From week 18 onwards, the selected males and all females were transferred to the layer house and kept in floor pens with 1 cock and 10 hens per pen. Pens were fitted with trap nests to facilitate full pedigree recording. Eggs were collected from selected hens for 10-12 days and incubated in three hatches to produce the next generation (Woldegiorgiss, 2015).Blood samples were collected from 27 chicken populations in Ethiopia (Figure 16). Samples included 103 cocks and 157 hens. Except the improved Horro, Meseret, Tsion Teguaz, Jarso and local Horro populations, 10 chicken from each village were sampled. One or two chicken were sampled per household. Improved Horro was sampled from a breeding stock of 8 th generation under selection at Debre Zeit Agricultural Research Centre and used as a reference population.Unlike other populations, Jarso and Local Horro sequences were obtained and included from the previous studies. Photographs and weight of each bird were taken. The average weight of sampled chicken was 1.26 kg with age ranges of 5 to 36 months. Sampling considered different agroecological zones, altitudes ranging from 729-3500 meters, marketing points, and chicken phenotypic characteristics. From the wing vein of each chicken, 50 -250 µl of whole blood were drawn with syringes using cryotubes filled with 1.5 ml absolute ethanol (100%) following the guidelines available at https://www.sheffield.ac.uk/nbaf-s/protocols_list.Total DNA was extracted from chicken whole blood at the BecA-ILRI Hub, Nairobi, Kenya facility (http://hub.africabiosciences.org/) using the Qiagen DNeasy blood and tissue kit protocol (Lwelamira et al., 2008). To evaluate the DNA concentration a Thermo Scientific NanoDrop spectrophotometer 2000c was used. The integrity of DNA was confirmed by agarose gel electrophoresis whereby 20 ng/µl genomic DNA samples were loaded with 1 µl loading dye (6X) on a 1% agarose gel containing 2.5 µl gel red at a voltage of 7/cm for 60 minutes, 3 µl of lambda DNA of size of 48,500 bp and a concentration of 20 ng/µl was used as size marker and the gel was then examined using UV light using GelDoc-It2 Imager to check the extracted DNA quality and quantity. The genomic DNA from (n = 284) was normalized to a final volume of 100 µl and final concentration of 50 ng/µl and sent to Edinburgh Genomics, UK, for whole genome sequencing.The libraries were evaluated for mean peak size and quantity using the Caliper GX Touch with a HT DNA 1k/12K/HI SENS LabChip and HT DNA HI SENS Reagent Kit. Those libraries were then normalized to 5 nM using the GX data and the actual concentration was established using a Roche LightCycler 480 and a Kapa Illumina Library Quantification kit and Standards. The normalized libraries were denatured and pooled in eights for clustering and sequencing using aHamilton MicroLab STAR with Genologics Clarity LIMS X Edition. Libraries were clustered onto HiSeqX Flow cell v2.5 on cBot2s and the clustered flow cell is transferred to a HiSeqX for sequencing using a HiSeqX Ten Reagent kit v2.5. The samples were sequenced at a genome coverage of ~5-90X (mean = 36.1X). Demultiplexing is performed using bcl2fastq (2.17.1.14), allowing 1 mismatch when assigning reads to barcodes.Adapters (Read1: AGATCGGAAGAGCACACGTCTGAACTCCAGTCA, Read2:AGATCGGAAGAGCGTCGTGTAGGGAAAGAGTGT) are trimmed during the demultiplexing process.The pipelines for mapping and variant calling included: mapping reads against reference genome using BWA-mem, sorting BAM file, removing duplicated reads with PICARD, Base Quality Score Recalibration (BQSR) with GATK, calling variants using GATK, Variant Quality Score Recalibration (VQSR) for variant filtration with GATK, and finally selection of only bi-allelic SNPs which passed the VQSR step.High quality paired-end reads (FASTAQ format) were aligned to the chicken (Gallus gallus) Variant calling from each sample was performed in the gVCF mode for cohort analysis using GATK's HaplotypeCaller (Figure 2). Joint genotyping of samples from each population were done using GATK's GenotypeGVCF tool for downstream analysis. Variant Quality Score Recalibration (VQSR) were also performed to increase sensitivity (identifying the real variants) and specificity (identifying false positives) using GATK followed by a selection of only bi-allelic SNPs that passed the VQSR step. For the VQSR step, we used 1M validated SNPs and 15 SNPs from dbSNP for recalibration purpose.A total of 14, 857, 039 recalibrated autosomal SNPs were generated from 70 chicken samples (5 populations) and used for downstream analysis of signatures of selection. For each indigenous population we also produce environmental suitability maps for indigenous chicken populations based on the following environmental variables: Minimum temperature of the coldest month, precipitation seasonality, precipitation of the wettest quarter, precipitation of the driest quarter, % of cultivated land, % of grass/scrub/woodland, proportion of crop rainfed or irrigated and carbon content (g/kg).Selection sweep detection was carried out using Hp and Fst statistics using VCFtools version 0.1.13 in an overlapping window bin size of 20 kb and a step size of 10 kb. These statistics involves comparing the average number of nucleotide differences from pair wise DNA sequences and the number of segregating sites. Using the pool heterozygosity (Hp) method (Rubin et al., 2010), the levels of heterozygosity were measured for the autosomal genome (chromosomes 1-28 and 30-33) at a window of 20 kb and 10 kb step size.The pooled heterozygosity (Hp) values were calculated using the following equation:Where ∑ n MAJ and ∑ n MIN ) are the sums of major and minor allele frequencies respectively for all the SNPs within each the 20-kb window. At each detected SNP position, we counted the number of reads corresponding to the most and least frequently observed allele (nMAJ and nMIN, respectively) for each population. The values of Hp calculated for each window size were then subsequently Z-transformed using the equation:Where \uD835\uDC4B ̅ the mean and σ is is the standard deviation of the Hp value. Windows with a large number of heterozygote SNPs show values above zero, they may reflect balancing selection signature.Only windows with at least 20 SNPs were extracted and set for analysis. Following this criterion, 1117, 1152, 1029, 1444 and 911 windows with SNPs < 20 were excluded for Improved Horro,Local Horro, Arabo, Hugub and Jarso populations. Respectively. From the remaining windows, a genome-wide significant threshold score of Z (Hp) ≤ -4.0 was considered (Rubin et al., 2010).Population differentiation values (Fst) which compare differences in allele frequencies between population were calculated for each SNP as described in Akey et al. (2002). Fst was calculated from the allele frequencies (not the allele counts) using the standard equation:Where, P i within = P(i) population 1+ P(i ) population 2 2andwith fN being the frequency of nucleotide N (A, T, C or G), P i total is the total P i for which allele frequencies in both populations are averaged. The Fst values were Z-transformed as follows:Where, μ is the mean and σ is the standard deviation of the Fst.Putatively selected regions were selected based on windows within the 1% level low and high ZFst values.To establish the biological significance of the genes found within each candidate selected region, the genes putatively under selection were submitted to DAVID Bioinformatics Resources 6.8(https://david.ncifcrf.gov/) for enrichment analysis of the Gene Ontology (GO) and the Kyoto Encyclopaedia of Genes and Genomes (KEGG) pathways (KOBAS version 3.0, http://kobas.cbi.pku.edu.cn/). All chicken genes annotated in Ensembl were included. The two analyses restricted over-represented genes to the Fisher exact P-value < 0.05 default threshold.The genetic structure of the populations was examined on the basis of all SNPs using principal component analysis (PCA) (Figure 30). The first two components account for 19.27% and 7.1% of the variation respectively. They separate clearly the five populations. In Improved Horro chicken population, from the total number of 91,996 windows 90,877 windows with 20 and above SNPs were analyzed (Table S 48). The mean Hp value is 0.33 ± 0.058, while the minimum and maximum ZHp is -5.71 and 2.88, respectively, with 417 windows below the genome-wide threshold of ≤ -4 (Table S 48). The proportion of windows with significant ZHp in Improved Horro is 0.46% (417*100)/90878). These windows define 417 candidate sweep regions including 125 annotated genes (Table S 49). Chromosomes 1 to 15, chromosome 17, chromosome 20 and Chromosome 24 show significant window ≤ -4 (Figure 32). Across the genome, the strongest peak is located on chromosome 1 and 12 (Chr1: 189490000 to 189510000 bp; Chr12: 190890000-190930000 bp regions) with a ZHp score of -5.71 for the chromosome 1, including the Thyroid Hormone Stimulating Receptor (THSR) gene (Table 20).3.  133In Local Horro chicken population, from a total of 91905 windows, 90753 windows with 20 and above the number of SNPs were analyzed of which 311 windows (0.34%) passed the genomewide significant threshold of ≤ -4 (ZHp) (Table S 50). These define 311 candidate sweep regions including 83 genes (Table S 51). Chromosomes 1 to 15, chromosome 17, chromosome 20 and chromosome 24 show significant peaks (Figure 34). Across the genome, the strongest peaks are   In Jarso chicken population, from a total number of 91966 windows, 91055 windows with 20 and above the number of SNPs were analyzed (Table S 52) of which, 31 windows (0.034%) passed the genome-wide significant threshold of < -4 windows (ZHp). Chromosomes 1 to 9 and chromosome 13 have significant peaks. Thirteen annotated genes are found in these windows (Table S 53).Across the Jarso genome, the strongest peak is located on chromosome 13 (position 520,000 to 550000 bp) with a ZHp score of -4.68 (Figure 36). In Hugub chicken population, from the total number of 92006 windows, 90560 windows with 20 and above number or SNPs were analyzed (Table S 54) of which, 242 windows (0.12%) passed the genome-wide significant threshold of ≤ -4 (ZHp) (Table S 55). The mean Hp value for the window across the genome is 0.33 ± 0.063. Whereas the minimum and maximum Z-transformed pooled heterozygosity value are -5.12 and 2.72, respectively. These significant regions include 78 annotated genes (Table S 56). Chromosomes 1 to 9 and chromosome 13 show significant peaks ≤ -4 (Figure 38). Across the genome, the strongest peaks are observed on chromosome 3 (50630000-50650000; 7020000-7040000; 7930000 -7960000; 18930000-18950000 bp) with a ZHp score of -5.12. A gene of particular biological interest on chromosome 3 (position: 26395277 to 26573746)is the Protein Kinase C Epsilon (PRKCE).  In Arabo chicken population, from the total number of 92250 windows, 91221 windows with 20 and above number SNPs were analyzed (Table S 57). The mean Hp value is 0.30 ± 0.061, while the Z transformed minimum and maximum Hp value are -4.896 and 3.02, respectively. Fifty windows (0.055%) passed the genome-wide significant threshold of ≤ -4 (ZHp) (Table S 57).These windows defined 50 candidate sweep regions with 14 annotated genes (Table S 58).Chromosomes 1 to 9 and Chromosome 13 show significant peaks. Across the Arabo chicken genome, the strongest peak is located on chromosome 3 (82480000-82500000bp) (Figure 40    Stop gain/lost 0(2) 4( 1) 2(0)Micro RNA 1 Upstream: a variant that is located in the 1-kb region upstream of the gene start site; stop gain: a non-synonymous (ns) SNP that leads to the creation of a stop codon at the variant site; stop loss: a non-synonymous SNP that leads to the elimination of a stop codon at the variant site; splicing: a variant within 2 bp of a splice junction; downstream: a variant that is located in the 1-kb region downstream of the gene end site; upstream/downstream: a variant that is located in the downstream and upstream regions of two genes. The genome of the considered chicken populations was checked for overlapping sweep regions.Sixty-four windows were merged and checked for duplicate/overlapped regions (Table S 67). 145 duplicate values were obtained from the regions across populations. Finally, 64 regions were found in overlapping regions between chicken populations. From these regions, 31 genes were obtained for further functional annotation (Table S 68). Enriched functions for commonly selected regions include calcium signaling and other biological processes (Table 20).  23).Pair wise Fst results between Improved Horro and other indigenous chicken populations of Ethiopia namely Local Horro, Jarso, Hugub, and Arabo chicken populations are given below.From the total pair wise Fst windows (91,989) between Improved and Local Horro 90,572windows with greater than 20 SNPs, the top 1% windows (906) were considered as significant windows for downstream analysis (Table S 59). The minimum and maximum Fst values were 0.04 and 0.15, respectively. The significant genetic differentiation based on Fst values was mainly concentrated in the majority of the chromosomes Table 24). Detection of selection at the genome level using the Fst outlier method yield 311 candidate genes showing high evidence of positive selection from these significant regions (Table S 60). Between the highly differentiated regions of these populations, an interesting gene called Ral GTPase activating protein catalytic alpha subunit 1 (RALGAPA1) has been found.   The highest Fst variant statistic were obtained (n = 157,323) between Improved Horro and Local Horro followed by Improved Horro and Jarso (n = 135,913) (Table 21. The highest proportion of novel Fst variants are between Improved Horro and Jarso (14.6%) followed by Improved Horro and Arabo (13.2%). The highest mean Fst (0.09 ± 0.02) is between IH and LH but the maximum genetic differentiation was between IH and HUG (Fst = 0.82). In terms of Fst variant consequences, the highest missense variant is obtained between IH and LH followed by IH VS HU (Table 22). The number of Fst variants based on SIFT prediction is indicated in Table 23. The highest deleterious variants are found between IH and LH (n = 180) followed by IH and AR (n = 127). Splice donor/acceptor 10( 9) 21( 4) 20( 19) 10( 5)Stop gain/lost 10(2) 5( 4) 12( 7) 5( 1)Start lost / regained 5(4) 4 IH = Improved Horro; LH = local Horro; JR = Jarso; Hu = Hugub; 5UTR = 5 prime untranslated region; 3UTR = 3 prime untranslated region Upstream: a variant that is located in the 1-kb region upstream of the gene start site; stop gain: a non-synonymous (ns) SNP that leads to the creation of a stop codon at the variant site; stop loss: a non-synonymous SNP that leads to the elimination of a stop codon at the variant site; splicing: a variant within 2 bp of a splice junction; downstream: a variant that is located in the 1-kb region downstream of the gene end site; upstream/downstream: a variant that is located in the downstream and upstream regions of two genes. The selective sweep can have a dramatic impact on the level of population subdivision, particularly when the sweep has not yet spread to all populations within a species (Nielsen et al., 2011). The size of a selective sweep may depend on factors such as the local recombination rate, whether the selected variant ever reached complete fixation, the number of generations it took before fixation and any population admixture at a time point after the sweep initially occurred (Rubin et al., 2010).Across the Improved Horro genome, the strongest peak is located on chromosome 1 and 12(189490000 to 189510000 bp) with a Z Hp score of -5.71 including Thyroid Stimulating HormoneReceptor (TSHR) and other variable genes. The TSHR gene (Chromosome 5: 40811286-40858950 bp) which is a previously reported locus with a pivotal role in metabolic regulation and reproduction process (Rubin et al., 2010), is reported in Improved Horro chicken population. It is regarded as one of the most striking selective sweeps found in all domestic chicken. The established selective sweeps around the TSHR gene in domestic chicken is identified in the five chicken populations and this was considered as a proof of principle demonstrated that the identification of selection signals using Hp methods is reliable. Besides, the previously reported gene General Transcription Factor IIA Subunit 1 (GTF2A1) (Chromosome5: 40868271-40894704 bp) known to be involved in the production of eggs in birds is also under strong selection pressure in this population ( Lawal et al., 2018;Yuan et al., 2015).Another interesting phenomenon is the presence of Myozenin 1 (MYOZ1) gene is also found to be under the pressure of the ongoing selection program. Myoz1 gene under the candidate signatures of selection which plays a crucial role in signal transduction and muscle fiber type differentiation. The Myoz1 gene is a potential candidate for affecting carcass and meat quality traits in animals (Luo et al., 2018). Among other numerous genes in this study, the previously reported ovostatin (OVST) gene which is associated with the formation of eggshells by regulating eggshell matrix protein secretion is also under strong selection (Cordeiro and Hincke, 2016).Angiotensin II Type 1 Receptor (AGTR1) (Chromosome 9: 12398415-12430615 bp) gene is the other gene available in the selective sweeps of this specific population. This gene is also reported heavily involved in Ascites in commercial broilers (Krishnamoorthy et al., 2014). Ascites refers to abnormal accumulation fluid in the abdominal (peritoneal) cavity and it is a disease of modern days in the poultry industry (Qanbari et al., 2015b;Wideman et al., 2013). In humans, AGTR1(Chromosome 9: 12398415-12430615 bp) is a strong candidate for the pulmonary arterial hypertension (Burks, 2011;Chung et al., 2014;Crossley and Altimiras, 2012). The Immunoglobin Superfamily Member 21 (IGSF21) (Chromosome 21: 23092-65640 bp) which promotes differentiation of inhibitory synapses via binding to neurexin2α is also under selection pressure (Tanabe et al., 2017). Coordinated development of excitatory and inhibitory synapses is essential for higher brain function, and impairment in this development is associated with neuropsychiatric disorders. The most selected chromosome in Improved Horro is chromosome 1. Genes of interest that contain statistically significant include the DnaJ heat shock protein family (Hsp40) member C12 (DNAJC12) (Chromosome 6: 6666081-6675111 bp) in Improved Horro chicken population.The DNAJC12 gene plays a pivotal role in the negative regulation of neuron apoptotic process (Fleming et al., 2017).On top of TSHR, GTF2A1, AGTR1, and many other genes, the Beta-Carotene Dioxygenase 2 (BCDO2) (Chromosome24: 6130965 -6110301) gene is the only gene uniquely available in the candidate signature of selection in Local Horro. This gene is known to express in the skin where it encodes an enzyme that cleaves colorful carotenoids into colorless apocarotenoids, and polymorphisms in the BCDO2 gene have well-known effects on skin pigmentation in birds (Eriksson et al., 2008). To this end, it looks that the ongoing Improved Horro selection and improvement operation is working against this specific gene as it is not found in the selection signature regions of Improved Horro.Unlike other populations, Jarso chicken populations have a fewer number of variants under strong selection of signature. In addition to the common genes TSHR and GTF2A1, many other genes Signal transducer and activator of transcription 5b (STAT5b), the gene responsible for bone allocation and fecundity trait is found between Improved and Local Horro (Fu et al., 2016). Signal transducer and activator of transcription 5b (STAT5b) gene is found candidate gene and it is associated with body weight and reproductive traits of Jinghai Yellow chicken (Luo et al., 2018;Zhou et al., 2005).The only gene found in significant regions between these two populations is a gene called AGTR1(Angiotensin II (Ang II)) which is an important regulator of cardiovascular function in adult vertebrates and have roles in thermoregulation (Crossley et al., 2010). This gene is known to heavily involve in Ascites in commercial broilers (Krishnamoorthy et al., 2014). The other gene is roundabout guidance receptor 2 (ROBO2) (96895939-97053385 bp) gene which belongs to the immunoglobulin superfamily and plays functions associated in axon guidance and cell migration and are involved in SLIT/ROBO signaling (Wang et al., 2014). The ROBO2 gene has a strong effect on the antibody response to the NDV in chickens (Luo et al., 2013). The RALGAPA1 (36275772-36390043 bp) gene which is known to play a pivotal role in reproductive traits and broodiness is also under strong selection in Improved Horro and Local Horro chicken populations (Shen et al., 2012).A gene called unconventional myosin-VI; MYO6 (80736607-80807004 bp) which serve in intracellular movements are also found between the high confidence selection regions of Improved Horro and Jarso chicken populations. Myosin 6 is a reverse-direction motor protein that moves towards the minus-end of actin filaments. The gene of interest is Ankyrin 2 (ANK2) (57097275-57432432 bp) which was reported by Fan et al. (2013) previously which Ankyrins play key roles in activities such as cell motility, activation, proliferation, contact and the maintenance of specialized membrane domains. Like the ROBO2 gene, the ROBO1 gene is known to have a strong effect on the antibody response to the NDV in chicken (Luo et al., 2013). Another interesting gene in these populations is Interleukin-15 (IL-15) are T-cell growth factors potentially capable of enhancing cell-mediated immunity in vivo and plays a critical role in immune system function (Lillehoj et al., 2001)  (Kannaki et al., 2010). In response to pathogen-associated molecular patterns, TLRs induce the production of reactive oxygen and nitrogen intermediates, inflammatory cytokines and upregulate the expression of co-stimulatory molecules, subsequently initiating adaptive immunity (Ibid).Among other genes, the Insulin-like Growth Factor 1 Receptor (IGF1R) and Insulin-like Growth factor 2 mRNA-Binding Protein 3 (IGF2BP3) genes, which are necessary for formal growth (Rubin et al., 2010;Stainton et al., 2015;Yang et al., 2014), is found between Improved Horro and Arabo chicken populations. Between these populations, the myosin-binding protein C, cardiac-type (MYBPC3) and GATA binding protein 3 (GATA3) genes are under strong pressure of selection. MYBPC3 gene is known as an accessory protein of vertebrate striated muscle thick filaments that modulate cardiac muscle contraction (Carrier et al., 2015). Haploinsufficiency for the transcription factor GATA3 leads to hearing the loss in humans. It is expressed throughout the auditory sensory epithelium (SE) (Alvarado, 2009;Alvarado et al., 2009). Integrin alpha-8/beta-1 (ITGA8) functions in the genesis of kidney and probably of other organs by regulating the recruitment of mesenchymal cells into epithelial structures. Candidate Fst signals between IH and Arabo population also evidences the presence of stress-related genes, Hypocretin (orexin) neuropeptide (HCRT) (Fleming et al., 2017) in the candidate signature of the selection region. The previously reported genes, TBC1 domain family member 7 (TBC1D7) and TBC1 domain family member 30 (TBC1D30) genes which are associated with hypothermia and stress are found in this population (Fleming et al., 2017).Since most of the candidate genes identified in the present study are novel and have probably been Characterization of farm animal genetic resources is a prerequisite for any improvement, conservation and sustainable utilization of these resources (Al-Qamashoui et al., 2014). Accordingto Kristensen et al. (2015), genetic improvement of animals is dependent on the existence of genetic variation existing between species, between breeds, within species and among animals within breeds. The genetic diversity comprised in farm animal species and breed is an important resource in livestock systems (Oldenbroek, 2007). However, as species and breeds are adapted to certain environments through centuries of natural and artificial selection, it may be difficult to restore genetic variation that may still be desired, but that has been lost by breed replacements in certain regions or environments (Kristensen et al., 2015). These days, global climate change is increasing the magnitude of environmental stressors, such as temperature, pathogens, and drought that limit the survivability and sustainability of livestock production (Fleming et al., 2017). This scenario has demanded the poultry industry to heavily depend upon robust animals that are able to cope with multiple environmental stressors. In this regard, indigenous chickens are endowed with high genetic variation and high local adaptation caliber which forms the basis of selective breeding and genetic improvement strategies through prioritizing and making informed decisions (Desta, 2015;Lawal et al., 2018).Besides, to meet the future food demand and ensure plane of family nutrition in low input poultry systems, it is important to improve the productivity of indigenous chickens (Al-Qamashoui et al., 2014). Chicken is a major protein source and intensively selected for economically important traits by humans (Desta, 2015). However, for a multitude of facts within the different species used for food production, only a few breeds are developed towards high-output breed fitting in high input systems leaving aside enormous breeds from the food producing livestock systems and exposing to high danger of extinction (Oldenbroek, 2007). Indigenous chicken is endowed with a multitude of important traits with a high reputation for hardiness and resistance to diseases which needs to be better characterized for better utilization and conservation (Ngeno et al., 2014). Years of natural selection, under scavenging conditions, has made them robust and resistant to various diseases, especially to those caused by bacteria, and protozoa and other internal and external parasites; they have better survival than the commercial hybrid strains under village production conditions (Besbes et al., n.d.). The village chicken is very alert and has long shanks to run away from predators and multiple colors to serve as camouflage against aerial predators. Indigenous chickens appear to have an inherent scavenging and nesting habit.Although, purebred selection in indigenous chicken (IC) population require more time to improve performance than crossbreeding and breed substitution, nevertheless, it can be tailored to fit the needs of local farmers and the prevailing environmental conditions (Ngeno et al., 2015). Besbes Ethiopia and has resulted in better egg production, compared to unimproved village chicken (Woldegiorgiss, 2015;Wondmeneh et al., 2016). Ethiopian Horro IC has increased egg production by 123.5% (75 eggs) by week 45 and age at first egg reduced to 148 from 203 days by generation five in five generations of selection for egg numbers.To date, few and non-exhaustive studies have been carried out to characterize and unravel the genomic potential of indigenous chicken in Ethiopia using the state of the art technologies.Understanding the phenotypic and genomic diversity is a prerequisite for proper utilization and improvement of indigenous chicken in Ethiopia. Understanding the functional basis of the genetic variants that underlie these traits, however, remains a formidable endeavour particularly for complex traits. Nonetheless, molecular phenotyping of an organism from sequenced data is doable with the advances in bioinformatics analysis and unparalleled surveys of genome-wide genetic variants (Khoo, 2017). The aim of this study was therefore to undertake genome characterization of indigenous chicken in Ethiopia. The findings in this thesis provide knowledge on the genomic diversity, and regions under positive selection pressure adapted to different agro-ecological environments in different geographically distributed indigenous chicken in Ethiopia.In chapter 1, we have tried to emphasize the pinning problems that justify this work and gives backgrounds and objectives. The second chapter comprehensively revises basic concepts of chicken phenotypic and genome diversity and the molecular methods to tap these variations. In line with the previous studies (Alemayhu, 2003;Bekerie, 2015;Desta et al., 2013;Hassen et al., 2009;Lawal et al., 2018;Mwacharo et al., 2013Mwacharo et al., , 2007;;Wragg et al., 2012) strong candidates for growth loci in the chicken using linkage mapping approach to map growth traits in an advanced intercross of wild red jungle fowl and domestic white leghorn layer chickens.Of which only the gene, CEP55, has been found in our list of genes as non-synonymous deleterious variants across 27 populations. Our informed analysis of a previously reported gene, Neural crest Hypothesis Domestication (FGFR-I) and Gonadotrophin-releasing hormone I (GnRH-I) doesn't ascertain its presence in the candidates of signature of selection in Improved Horro and other indigenous chicken populations of Ethiopia. In A similar fasion, the vasoactive intestinal polypeptide receptor-1 (VIPR-1) and dopamine D2 receptor (DRD2) genes reported by Xu et al (2011) as having an association to chicken egg number at 300 days of age is reported.In • Future works should emphasize to map candidate genes responsible for egg and meat production.• Chicken improvement programs ahead should consider the 4 gene pools revealed in this study.• Various characterization studies have been done here and there on a piecemeal basis. Hence, concerted efforts should gear towards Genetic improvement programs through selection.• Using the current data it is also important to check if the regions under the current selection pressure overlap to the Quantitative Trait Loci (QTLs) available on the online animal QTL data base.• Sequence data management of this finding was based on Gal gal 5.0 version and thus future works should also consider re-analysis based on the upcoming version Gal gal 6.0 which is not yet publicized for use.            Dana, 2010;Desta et al., 2013;Moges et al., 2010;Getu et al., 2014aGetu et al., , 2014b;;Hailu, et al., 2013;Halima et al., 2007;Kibret, 2008;Moges, 2014;Lemlem and Tesfaye, 2010;Negassa et al., 2014;Yitbarek and Zewudu;Zewdu, 201316 Alemayhu, 2003;Bekerie, 2015;Dana, 2010;Goraga et al., 2012;Hassen et al., 2009;Mwacharo et al., 2007;Desta et al., 2014;Wragg et al., 2012 8 N = number of studies Phenotypic studies N Genotypic studies N 500 -1800 Alemayhu, 2003;Bekele et al., 2015;Bogale, 2011;Dana et al., 2010;Duguma, 2006;Getachew et al., 2016;Halima et al., 2007;Nigussie, 2013;Yitbarek and Zewudu, 2013;Yisma, 2015;Zewdu et al., 2013. 10 Alemayhu, 2003;Hassen et al., 2009;Wragg et al., 2012. 4 1800-2400Aklilu, 2013;Alemayhu, 2003;Bekerie, 2015;Bikila, 2013;Dana, 2010;Duguma, 2006;Getachew et al., 2016;Getu et al., 2014b;Hailu et al., 2013;Halima, 2007;Kibret, 2008;Lemlem and Tesfaye, 2010;Mogess et al., 2014;Negassa et al., 2014;Nigussie, 2013;Yisma, 2015. 18 Alemayhu, 2003;Goraga et al., 2012a;Hassen et al., 2009;Mwacharo et al., 2007;Desta et al., 2014;Wragg et al., 2012. 6 2400-3200 Aklilu, 2013;Bekerie, 2015;Dana et al., 2010;Duguma et al., 2006;Getu et al., 2014b;Halima, 2007;Mogess, 2014;Nigussie, 2013;Dessie et al., 2013;Yisma, 2015. 10 Alemayhu, 2003;Bekerie, 2015;Mwacharo et al., 2007;Desta et al., 2014;Wragg et al., 2012.   et al., 2015;Bekerie, 2015;Dana, 2011a;Duguma, 2006;Getachew et al., 2016;Getu et al., 2014aGetu et al., , 2014b;;Halima et al., 2007;Kibret, 2008;Melesse and Negesse, 2011;Negassa et al., 2014;Nigussie, 2013;Yisma, 2015. 2 Comb type Aklilu, 2013;Bekele et al., 2015;Bekerie, 2015;Dana, 2011a;Desta et al., 2013;Duguma, 2006;Getu et al., 2014aGetu et al., , 2014b;;Halima et al., 2007;Kibret, 2008;Melesse and Negesse, 2011;Nigussie, 2013;Yisma, 2015, Getachew et al., 2016;Negassa et al., 2014. 3 Earlobe colour Aklilu, 2013;Bekele et al., 2015;Bekerie, 2015;Dana, 2011a;Desta et al., 2013;Duguma, 2006;Getu et al., 2014aGetu et al., , 2014b;;Kibret, 2008;Melesse and Negesse, 2011;Negassa et al., 2014;Nigussie, 2013;Yisma, 2015. 4 Eye colour Aklilu, 2013;Duguma, 2006;Getu et al., 2014aGetu et al., , 2014b;;Negassa et al., 2014;Yisma, 2015. 5 Shank colour Aklilu, 2013;Bekele et al., 2015;Bekerie, 2015;Dana, 2011a;Desta et al., 2013;Duguma, 2006;Getu et al., 2014b;Halima et al., 2007;Kibret, 2008;Melesse and Negesse, 2011;Negassa et al., 2014;Yisma, 2015. 6 Skin colour Aklilu, 2013;Bekele et al., 2015;Dana, 2011a;Duguma, 2006;Getu et al., 2014;Kibret, 2008;Negassa et al.,    , 2010;Aklilu, 2013;Bekele et al., 2015;Bekerie, 2015;Dana, 2011a;Duguma, 2006;Getachew et al., 2015;Getu et al., 2014aGetu et al., , 2014b;;Halima et al., 2007;Kibret, 2008;Melesse and Negesse, 2011;Negassa et al., 2014;Nigussie, 2013;Yisma, 2015. Egg/hen/year 15 Lemlem and Tesfaye, 2010;Aklilu, 2013;Bekele et al., 2015;Bekerie, 2015;Dana, 2011a;Duguma, 2006;Getachew et al., 2015;Getu et al., 2014aGetu et al., , 2014b;;Halima et al., 2007;Kibret, 2008;Melesse and Negesse, 2011;Negassa et al., 2014;Nigussie, 2013   3 Goraga et al., 2012;Mwacharo et al., 2007;Wragg et al., 2012. 5 Observed Heterozygosity 6 Alemayhu, 2003;Bekerie, 2015;Goraga et al., 2012;Halima et al., 2007;Mwacharo et al., 2007;Wragg et al., 2012. 6 Expected heterozygosity 7 Alemayhu, 2003;Bekerie, 2015;Goraga et al., 2012;Halima et al., 2007;Mwacharo et al., 2007;Desta et al.,         et al., 2015;Bekerie, 2015;Dana et al., 2010;Desta et al., 2013;Getachew et al., 2016;Getu et al., 2014aGetu et al., , 2014b;;Halima et al., 2007;Kibret, 2008;Melesse and Negesse, 2011;Negassa et al., 2014;Nigussie et al., 2015;Yisma, 2015) Rose 59 30.79 Bekele et al., 2015;Bekerie, 2015;Dana et al., 2010;Desta et al., 2013;Getachew et al., 2016;Getu et al., 2014aGetu et al., , 2014b;;Halima et al., 2007;Kibret, 2008                                   "}
\ No newline at end of file