LimmaRP

LimmaRP identifies differentially regulated features in high-dimensional biological datasets with limited replicates and substantial missing values, particularly for mass spectrometry-based proteomics and phospho-proteomics.

Key Features:

• Handling of Missing Values: Optimized to analyze datasets with a high proportion of missing values, including mass spectrometry-driven proteomics and phospho-proteomics experiments.
• Statistical Methodology: Integrates standard t-tests, limma (moderated t-test), and an improved rank products method optimized for sparse data.
• Complementary Analysis: Combines limma and the enhanced rank products approach to maximize identification of differentially represented features in datasets with more than 1,000 features and over 50% missing values.

Scientific Applications:

• Proteomics: Detection of differentially represented proteins or peptides from stable isotope labeling and mass spectrometry experiments with few replicates.
• Phospho-proteomics: Identification of significant peptide-level changes in phospho-proteomics datasets despite substantial missing values.

Methodology:

Performance was assessed using simulated and experimental datasets with varying degrees of missing data; statistical detection employs standard t-tests, moderated t-tests via limma, and an enhanced rank products method; an R script implements the improved rank products algorithm alongside the combined analysis approach.