ROCS: a reproducibility index and confidence score for interaction proteomics studies

Area of Science:

• Proteomics and Bioinformatics
• Molecular and Cellular Biology

Background:

• Affinity-Purification Mass-Spectrometry (AP-MS) is crucial for identifying protein complexes and interactions.
• AP-MS experiments face challenges with low reproducibility and high rates of false positives/negatives.
• Existing methods require improvements for accurate and reliable protein-protein interaction (PPI) identification.

Purpose of the Study:

• To introduce ROCS, a novel two-step method to enhance the accuracy and reproducibility of AP-MS data analysis.
• To address the limitations of low experimental reproducibility and inaccurate PPI identification in AP-MS studies.
• To provide an objective, error-controlled approach for analyzing AP-MS results.

Main Methods:

• ROCS utilizes Indicator Prey Proteins to select reproducible AP-MS experiments, filtering out noisy data.
• A Confidence Score is computed for prey proteins, assessing their probability of occurrence in bait versus control experiments.
• The method employs automatic objective criteria for parameter estimation and controls for false discovery rates and biological coherence.

Main Results:

• Application of ROCS to five AP-MS datasets demonstrated its ability to accurately identify specific, biologically relevant PPIs.
• ROCS can be used independently or in conjunction with other AP-MS scoring methods to improve inference quality.
• Systematic benchmarking confirmed the effectiveness of ROCS in enhancing AP-MS data interpretation.

Conclusions:

• ROCS is a valuable tool for addressing reproducibility and accuracy issues in AP-MS datasets.
• The methodology shows promise for broader application in proteomics studies and databases facing experimental variability.
• The ROCS method is implemented as a freely available R package on CRAN.