SePreSA

SePreSA predicts population-specific susceptibility to serious adverse drug reactions (SADRs) by integrating DOCK-based molecular docking, structural models, a 2-directional Z-transformation scoring algorithm, and population polymorphism annotations.

Key Features:

• Molecular docking (DOCK): Performs molecular docking of submitted drug molecules against multiple SADR protein targets using the DOCK program.
• Structural model library: Uses an extensive collection of structural models covering nearly all well-known SADR targets.
• 2-directional Z-transformation scoring: Applies a 2-directional Z-transformation scoring algorithm to compute relative drug–protein interaction strengths from the docking-score matrix within a chemical–protein interactome.
• Chemical–protein interactome (docking-score matrix): Constructs a docking-score matrix representing the chemical–protein interactome to organize interaction data.
• Target prioritization: Prioritizes SADR targets by ranking relative interaction strengths, aiming to improve accuracy relative to traditional docking scoring functions.
• 3D binding visualization: Generates 3D visualization of the binding pose with the lowest docking score and highlights amino acid residues involved in drug binding.
• Population polymorphism annotation: Annotates interacting residues with population-specific polymorphism information to enable inference of differential susceptibilities.

Scientific Applications:

• Population susceptibility prediction: Identification of populations susceptible to SADRs by integrating docking results with polymorphism annotations.
• Drug safety assessment: Supporting prioritization of SADR targets for experimental validation and risk mitigation in drug development.
• Target and variant selection: Informing selection of protein targets and genetic variants for follow-up studies and risk stratification.

Methodology:

Accepts a drug molecule input, performs DOCK-based molecular docking against structural models of SADR targets, compiles results into a docking-score matrix (chemical–protein interactome), applies a 2-directional Z-transformation scoring algorithm to compute relative interaction strengths and prioritize targets, generates 3D visualization of the lowest docking-score binding pose and highlights interacting amino acid residues, and annotates those residues with population-specific polymorphism information.