Oktoberfest: Open‐source spectral library generation and rescoring pipeline based on Prosit

• DOI: 10.1002/pmic.202300112
• Publication Date: 2023-09-06T21:38:34Z
• Authors: Mario Picciani; Wassim Gabriel; Victor‐George Giu...; Omar Shouman; Firas Hamood; Ludwig Lautenbacher; Cecilia Bang Jensen; Julian Müller; Mostafa Kalhor; Armin Soleymaniniya; Bernhard Kuster; Matthew The; Mathias Wilhelm
• Keywords: Python

Abstract Machine learning (ML) and deep (DL) models for peptide property prediction such as Prosit have enabled the creation of high quality in silico reference libraries. These libraries are used various applications, ranging from data‐independent acquisition (DIA) data analysis to data‐driven rescoring search engine results. Here, we present Oktoberfest, an open source Python package our spectral library generation pipeline originally only available online via ProteomicsDB. Oktoberfest is largely agnostic provides access predictions, promoting adoption state‐of‐the‐art ML/DL proteomics pipelines. We demonstrate its ability reproduce even improve results previously published analyses on two distinct use cases. freely GitHub ( https://github.com/wilhelm‐lab/oktoberfest ) can easily be installed locally through cross‐platform PyPI package.