RAPID

Support Vector Regression-Based Protein Disorder Estimation

RAPID (Regression-based Accurate Predictor of Intrinsic Disorder) estimates intrinsic disorder content across protein sequences using support vector regression and aggregated sequence-derived features.

Key Features:

• High Throughput Prediction: Predicts disorder content of an average-sized eukaryotic proteome in less than one hour on a modern desktop computer.
• Feature Integration: Integrates multiple complementary information sources aggregated over an input protein chain using advanced feature selection techniques.
• Competitive Performance: Demonstrates competitive performance against state-of-the-art disorder and disorder content predictors on diverse benchmark datasets.

Scientific Applications:

• Proteome-Scale Disorder Analysis: Characterizes intrinsic disorder across more than 200 fully sequenced eukaryotic proteomes and analyzes over 56,000 annotated human proteome chains to identify relationships between disorder content, structural coverage, chain length, fully disordered protein chains, and enrichment of cellular functions and localizations.

Methodology:

Applies support vector regression to aggregate multiple sequence-derived information sources through feature selection, enabling precise prediction of protein intrinsic disorder content.