iPNHOT

iPNHOT identifies hot spot residues at protein-nucleic acid interfaces to predict residues that contribute substantially to binding affinity in protein-DNA and protein-RNA complexes.

Key Features:

• Support Vector Machine (SVM): Employs a support vector machine classifier for hot-spot prediction.
• Atom Number of the Residue: Uses the atomic composition (atom number) of interacting residues as a predictive feature.
• Depth and Protrusion Index: Includes depth and protrusion indices to assess residue burial and exposure within the complex.
• Solvent Accessible Surface Area (SASA): Uses solvent accessible surface area to quantify residue exposure to solvent.
• Electrostatic Potential: Incorporates electrostatic potential to capture charge distribution around residues.
• Secondary Structure: Accounts for residue secondary structure context in the prediction.
• ∆SASsa1/2: Incorporates the novel ∆SASsa1/2 metric proposed in the study.
• esp3: Incorporates the novel esp3 electrostatic feature introduced in the study.
• Training Data: Trained on curated examples from the dbAMEPNI database and recent literature.
• Feature Selection: Applies a two-step feature selection strategy to identify seven key features for the final model.
• Input Format: Accepts protein-nucleic acid complexes provided in PDB format containing atomic coordinates.
• Performance Metrics: Reported performance on a ProNIT subset is F1 = 0.725 and AUROC = 0.807, compared to mCSM-NA (F1 = 0.407, AUROC = 0.670).

Scientific Applications:

• Hot-spot mapping: Map binding hot spots to elucidate binding mechanisms in transcription, translation, and gene regulation.
• Interaction characterization: Characterize residue contributions at protein-DNA and protein-RNA interfaces.
• Drug discovery: Inform design of molecules that modulate protein-nucleic acid interactions for therapeutic development.

Methodology:

Uses a support vector machine trained on data from dbAMEPNI and recent literature, with a two-step feature selection that selected seven features among atom number, depth and protrusion indices, SASA, electrostatic potential, secondary structure, ∆SASsa1/2 and esp3.