ProteinInfer

ProteinInfer computes conditional protein probabilities for protein inference in shotgun proteomics by applying combinatorial mathematics to peptide identification results under three assumptions (lower bound, upper bound, and empirical estimation).

Key Features:

• Combinatorial mathematics: Uses combinatorial mathematics to calculate conditional protein probabilities from peptide identification results.
• Conditional protein probabilities: Provides quantitative probabilities that reflect the likelihood that a given protein has been correctly identified.
• Three-assumption framework: Operates under three distinct assumptions providing a lower bound, an upper bound, and an empirical estimation of protein probabilities.
• Analytical expressions: Derives analytical expressions that relate protein identifications to peptide identifications.
• Peptide-type analysis: Evaluates the impact of unique peptides (those exclusive to a single protein) and degenerate peptides (shared by at least two proteins) on protein probabilities.
• Benchmarking: Benchmarked against ProteinProphet, reporting comparable accuracy and improved processing efficiency on standard protein mixtures and real sample datasets.

Scientific Applications:

• Protein inference in shotgun proteomics: Computes protein-level identifications and confidence estimates from peptide identification results in shotgun proteomics experiments.
• Peptide contribution analysis: Quantifies how unique and degenerate peptides influence protein identification probabilities.
• Method comparison: Enables comparative evaluation against ProteinProphet for method benchmarking.
• Dataset evaluation: Applied to standard protein mixtures and real biological sample datasets to assess inference performance.

Methodology:

Calculates conditional protein probabilities from peptide identification results using combinatorial mathematics under three explicit assumptions (lower bound, upper bound, empirical estimation), derives analytical expressions relating proteins and peptides, and performs comparative benchmarking against ProteinProphet.