Oktoberfest: Rescoring, Collision Energy Calibration and Spectral Library Generation for Proteomics

Oktoberfest is a python tool for collision energy calibration, rescoring search results and generating spectral libraries for proteomics research within the Prosit ecosystem. It offers an end to end pipeline that takes search results, predicts peptide properties using koina, plots summaries and quality control figures and performs FDR estimation with either mokapot or percolator.

Quickstart

Discuss usage, development and issues on GitHub.
Check our tutorial notebook, the usage principles or the API.
Check the Contributor Guide if you want to participate in developing.
Consider citing Oktoberfest along with original references.

News

Oktoberfest supports rescoring XL-MS datasets 2025-06-14

Oktoberfest 0.10.0 is published, introducing support for crosslinking mass spectrometry (XL-MS) rescoring — including both cleavable (DSSO, DSBU) and non-cleavable (DSS, BS3) crosslinkers.
The update adds compatibility with xiSEARCH output and integrates the new Prosit-XL models for XL-MS data.
Check out the publication in Nature Communications and explore the full tutorial notebook on GitHub.
Related config files can be found here: example_configs.py.
Oktoberfest is now also supporting python 3.12.

Oktoberfest supports OpenMS and custom modifications 2025-05-21

Oktoberfest 0.9.0 is published, allowing to provide custom modifications as long as they are supported by the respective Koina model. In addition, direct support of OpenMS searches was added.

This release supports python 3.10, and 3.11. Support for python 3.9 was dropped!

Oktoberfest supports requantifying rescored peptides with Picked-Group-FDR and offline predictions with custom models 2024-10-10

Oktoberfest 0.8.0 is published using Picked-Group-FDR, Oktoberfest can now requantify results from MaxQuant runs which enables downstream comparison of rescored with original results on protein level.
Additionally, custom features can now be added and configured to be used by percolator/mokapot for rescoring.
This release also adds preliminary support of local predictions via DLOmix. Stay tuned for a documentation update including jupyter notebooks.

Oktoberfest provides access to AlphaPept and MS2PIP predictions 2024-05-29

Oktoberfest 0.7.0 is published, providing access to predictions from AlphaPept and MS2PIP. This comes with a large API overhaul that allows easier access to underlying data for manual insights and analysis.
Please note that you can now specify the instrument type in the configuration for AlphaPept predictions.
Spectral libraries can now be generated using a simple peptide list using a variety of options, such as precursor charge, max. number of oxidations and collision energy, or directly in internal format. Check the new peptides / internal format specification and configuration options.
Stay tuned for a documentation update including jupyter notebooks on how to navigate and make use of your data for manual in-depth analysis.
This release drops python 3.8 support!

Support for timsTOF added / spectral library generation overhaul 2024-01-31

Oktoberfest 0.6.0 is out supporting the new Prosit model for timsTOF data. Check out the preprint at BioRxiv.
Spectral library generation was completely overhauled, resulting in more stability, dramatic decreases in runtime and storage requirements and more customization in the config.
Note that this release introduces changes in the configuration file that might break your previous setup! Please check the new configuration format.

Oktoberfest now provides a low level API 2023-10-03

With our new release 0.5.0, we refactored our code base to provide a low level API to the users. Check the new API section for more details.

Please note: We dropped support for percolator version 3.05. Please update to version 3.06.1. See installation for details.

Oktoberfest is published! 2023-09-07

Alongside the new 0.4.0 release, Oktoberfest was published in Proteomics.
Check out the publication here: https://doi.org/10.1002/pmic.202300112.
Accordingly, please check How to cite.

The first Oktoberfest release is ready! 2023-06-25

Version 0.3.0 is available on PyPI.

We are happy to announce that our first standalone version of Oktoberfest is finally ready. Besides a lot of bugfixes, new features and performance enhancements over the existing online Prosit service at ProteomicsDB, the Oktoberfest codebase is now open source and available on GitHub. In addition, we switched to our new online prediction service koina which allows connecting Oktoberfest to any public or self-hosted koina instance for easy prediction model integration and being independent of ProteomicsDB!