HBPred2.0

HBPred2.0 predicts soluble carrier hormone-binding proteins (HBPs) in humans and animals to facilitate identification of proteins involved in growth regulation and other biological processes.

Key Features:

• Target: Predicts soluble carrier hormone-binding proteins (HBPs) in human and animal proteomes.
• Machine Learning-Based Methodology: Employs a machine learning approach to distinguish HBPs from non-HBPs.
• Feature Encoding: Uses an optimal tripeptide composition encoding derived from a binomial distribution method.
• Performance: Demonstrated 97.15% overall accuracy in a 5-fold cross-validation assessment.

Scientific Applications:

• In silico identification: Enables computational identification of candidate HBPs to prioritize targets for experimental validation.
• Functional studies: Supports investigation of hormone–protein interactions and molecular mechanisms underlying growth regulation.
• Comparative proteomics: Can be applied to human and animal proteomes to compare HBP repertoires across species.

Methodology:

Machine learning approach using optimal tripeptide composition encoding derived from a binomial distribution method and evaluated by 5-fold cross-validation (97.15% overall accuracy).