Fig. 1: AlphaPept “ecosystem” and modules.
From: AlphaPept: a modern and open framework for MS-based proteomics
a AlphaPept relies on multiple community-tested packages. We use highly optimized libraries such as Numba, NumPy, CuPy, scikit-learn, SciPy, and pandas to achieve performant code. As GUI, we provide a browser-based application built on streamlit. For data handling, the HDF5 file technology is used. The repository itself is hosted on GitHub, and the core code is documented in Jupyter Notebooks using the nbdev package. To ensure maintainability, packages are continuously monitored for updates via dependabot. New code is automatically validated using GitHub actions and summary statistics (timing, identifications, and quantifications) are uploaded to a MongoDB database and visualized. b All algorithmic code of AlphaPept is organized in Jupyter Notebooks. For the key processing steps in the pipeline, such as importing raw data, Feature Finding, FASTA processing, Searching, Recalibrating, Scoring, Quantifying, and Matching, there are individual notebooks with background information and the code.
