H-VDW

H-VDW predicts quantitative hydrogen bond and van der Waals contact counts between protein and nucleic acid sequences to infer non-covalent interaction patterns from sequence data.

Key Features:

• Quantitative Prediction: Estimates numbers of hydrogen bonds and van der Waals contacts between protein and nucleic acid sequences.
• Hybrid Feature Set: Combines sequence-length fraction, conjoint triad descriptors for protein sequences, and gapped dinucleotide composition as input features.
• Amino Acid Polarity: Incorporates amino acid polarity as a biochemical feature influencing predicted hydrogen bond and van der Waals contacts.
• Support Vector Regression (SVR): Uses support vector regression models to generate quantitative interaction predictions from sequence-derived features.

Scientific Applications:

• Structural Prediction: Provides estimated non-covalent contact counts to assist modeling of protein–nucleic acid complex interactions from sequence data.
• Functional Analysis: Informs analysis of sequence-dependent effects on protein–nucleic acid binding by supplying predicted interaction patterns.
• Guiding Experimental Studies: Supplies preliminary interaction estimates to prioritize or guide experimental validation of protein–nucleic acid interactions.

Methodology:

Support vector regression models trained on sequence-derived hybrid features (sequence-length fraction, conjoint triad for proteins, gapped dinucleotide composition) with inclusion of amino acid polarity to predict hydrogen bonds and van der Waals contacts from sequences.