APRICOT

APRICOT identifies and characterizes RNA-binding proteins (RBPs) from protein sequences to enable sequence-based investigation of post-transcriptional gene regulation across eukaryotic species (including human and yeast) and bacterial systems.

Key Features:

• Sequence-Based Identification: Leverages known RNA-binding domains (RBDs) from experimental studies using position-specific scoring matrices and Hidden Markov Models to detect RBDs within protein sequences.
• Statistical Scoring: Applies a statistical framework that scores candidate RBPs based on sequence-derived features to distinguish true RBPs from non-binding proteins.
• Characterization of Putative RBPs: Analyzes multiple biological properties of predicted RBPs to provide functional insights.
• Adaptability and Performance: Validated on large-scale protein sets including the Escherichia coli proteome and demonstrated average sensitivity of 0.90 and specificity of 0.91, showing superior performance relative to existing sequence-based RBP prediction tools.

Scientific Applications:

• Post-Transcriptional Gene Regulation: Identifying and characterizing RBPs to inform studies of regulatory networks governing gene expression at the post-transcriptional level.
• Bacterial Proteomics: Enabling exploration of bacterial RBP diversity and functional roles in prokaryotic systems such as Escherichia coli.
• Comparative Genomics: Supporting comparative analyses across species using large protein datasets to discover conserved and species-specific RBP functions.

Methodology:

Uses known RNA-binding domains from experimental studies with position-specific scoring matrices and Hidden Markov Models to detect RBDs and employs a statistical scoring system on sequence-based features to prioritize putative RBPs.