Mass spectrometry-based proteomics data from thousands of HeLa control samples

Area of Science:

• Proteomics
• Machine Learning
• Biotechnology

Background:

• Mass spectrometry-based proteomics generates large datasets that are challenging to access and analyze.
• Researchers spend significant time filtering and preparing data for machine learning applications.

Purpose of the Study:

• To provide a curated, large-scale, label-free mass spectrometry proteomics dataset from HeLa cell lines.
• To facilitate machine learning and data analysis by offering machine-readable metadata and pre-filtered datasets.
• To enable automated data reprocessing through standardized instrument setting annotations.

Main Methods:

• Label-free mass spectrometry-based proteomics was performed on HeLa cell lines.
• MaxQuant software was used for data processing and analysis.
• Machine-readable metadata and standardized SDRF files were generated for all raw data.

Main Results:

• A comprehensive proteomics dataset comprising 7,444 raw files was generated.
• Machine-based metadata was created for efficient data selection and overview.
• Three aggregated development datasets (protein groups, peptides, precursors) were provided.
• SDRF files detailing instrument settings were included for automated reprocessing.

Conclusions:

• The provided dataset and associated tools significantly reduce the time and effort required for proteomics data analysis.
• This resource promotes broader adoption of machine learning in proteomics research.
• The open-access nature and provided workflows encourage community contribution and dataset expansion.