Systematic characterization and prediction of coenzyme A-associated proteins using sequence and network information

Area of Science:

• Biochemistry
• Bioinformatics
• Computational Biology

Background:

• Coenzyme A-associated proteins (CAPs) are vital for numerous biological functions via Coenzyme A (CoA) interactions.
• Systematic characterization and computational prediction methods for CAPs are currently lacking.

Purpose of the Study:

• To comprehensively characterize Coenzyme A-associated proteins (CAPs) and identify their unique features.
• To develop an accurate computational model for predicting CAPs across different species.

Main Methods:

• Multifaceted characterization of CAPs, including binding regions, evolutionary conservation, structural conformations, and network properties.
• Development of seven individual classifiers and a two-layer stacking model (CAPE) integrating various predictive features.
• Application of CAPE for high-confidence CAP identification and cross-species prediction.

Main Results:

• CAPs exhibit distinct characteristics, including CoA-binding regions, high evolutionary conservation, ordered/hydrophobic structures, and dense network localization.
• The developed classifiers and the CAPE model accurately identified CAPs in humans, mice, and Arabidopsis thaliana.
• CAPE demonstrated effective cross-species prediction capabilities, establishing a generic CAP prediction model.

Conclusions:

• This study provides a thorough survey of CAPs and introduces an effective computational predictor, CAPE.
• The findings facilitate the identification of novel CAPs and enhance understanding of the interplay between CoA and associated proteins.
• The developed generic model aids in uncovering the roles of CAPs in diverse biological contexts.