De novo peptide sequencing with InstaNovo

The official code repository for InstaNovo. This repo contains the code for training and inference of InstaNovo and InstaNovo+. InstaNovo is a transformer neural network with the ability to translate fragment ion peaks into the sequence of amino acids that make up the studied peptide(s). InstaNovo+, inspired by human intuition, is a multinomial diffusion model that further improves performance by iterative refinement of predicted sequences.

Publication in Nature Machine Intelligence: InstaNovo enables diffusion-powered de novo peptide sequencing in large-scale proteomics experiments

The full documentation is available at https://instadeepai.github.io/InstaNovo/ and consists of the following sections.

• Tutorials
  • How to install InstaNovo, make your first prediction and evaluate InstaNovo’s performance.
  • An end-to-end starter notebook that you can run in Google Colab .
• How-to guides:
  • How to perform predictions with InstaNovo with iterative refinement of InstaNovo+, or how to use each model separately.
  • Guide for preparing your own data for use with InstaNovo and InstaNovo+.
  • Details how to train your own InstaNovo and InstaNovo+ models.
• Reference:
  • Overview of the instanovo command-line interface.
  • List of the supported post translational modifications.
  • Description of the columns in the prediction output CSV
  • Code reference API
• Explanation:
  • Explains our performance metrics and benchmarking results
  • A detailed explanation of the SpectrumDataFrame class and its features.
• Blog:
  • Introducing the next generation of InstaNovo models
  • Introducing InstaNovo-P
  • Winnow: calibrated confidence and FDR control for de novo sequencing
• For Developers:
  • How to set up a development environment.
  • How to run the tests and lint the code.
  • View the test coverage and test report.
• How to Cite:
  • Bibtex references for our peer-reviewed publication on InstaNovo and InstaNovo+ and our preprints on InstaNovo-P, InstaNexus and Winnow.
• License:
  • Code is licensed under the Apache License, Version 2.0
  • The model checkpoints are licensed under Creative Commons Non-Commercial (CC BY-NC-SA 4.0)

Developed by:

• InstaDeep
• The Department of Biotechnology and Biomedicine - Technical University of Denmark