PROSE

PROSE computes interpretable enrichment scores from observed proteins using gene co-expression matrices to recover missing proteins and associate proteomic profiles with phenotypes.

Key Features:

• Scalability and Efficiency: Scalable, computationally efficient pipeline for processing large proteomic datasets.
• Enrichment Scoring: Computes interpretable and stable enrichment scores from an observed protein set that quantify each protein's relative importance to specific phenotypes.
• Missing Protein Recovery: Leverages gene co-expression matrices to reproducibly recover missing proteins from proteomic data, improving proteome coverage.
• Phenotype Association: Aligns enrichment scores with source phenotypes to relate protein presence or absence to phenotypic characteristics.

Scientific Applications:

• Proteomic dataset reanalysis (e.g., CCLE): Applied to reanalyze proteomic datasets such as the Cancer Cell Line Encyclopedia (CCLE) to enhance protein coverage and interpretation.
• Oncogenic dependency prediction: Used to predict oncogenic dependencies from proteomic profiles.
• Regulatory module identification: Used to identify well-defined regulatory modules from proteomic data.

Methodology:

Uses an input set of observed proteins and gene co-expression matrices to compute enrichment scores that quantify protein importance relative to phenotypes and enable recovery of missing proteins.