README.md

---
title: De Novo Peptide Sequencing With InstaNovo and InstaNovo+
emoji: 📊
colorFrom: green
colorTo: green
sdk: gradio
sdk_version: 5.23.1
app_file: app.py
pinned: true
license: apache-2.0
thumbnail: >-
  https://cdn-uploads.huggingface.co/production/uploads/6189aee17d9b289cdebafbd6/tb9e-8Z2_pDsRMkGglcvh.png
short_description: Translate fragment ion peaks into sequence of amino acids
---

# _De Novo_ Peptide Sequencing With InstaNovo and InstaNovo+

This Space provides a web interface for the [InstaNovo](https://github.com/instadeepai/InstaNovo) models for _de novo_ peptide sequencing from mass spectrometry data.

**Features:**

*   Upload MS/MS data in common formats (`.mgf`, `.mzml`, `.mzxml`).
*   Choose between fast Greedy Search or more accurate but slower Knapsack Beam Search.
*   View predictions directly in the interface.
*   Download full results as a CSV file.

**How to Use:**

1.  Upload your mass spectrometry data file.
2.  Select the desired decoding method.
3.  Click "Predict Sequences".
4.  View the results table and download the CSV if needed.

**Model:**

This demo uses the  pretrained model checkpoint.

* Predictions use version `instanovo-v1.1.0` for the transformer-based InstaNovo model and version `instanovoplus-v1.1.0` for the diffusion-based InstaNovo+ model.

**Note:** Processing large files can take time, depending on the file size and the chosen decoding method. Knapsack generation can also add to the initial startup time.