PharmoSB: A Direction-Aware Structure-Based Pharmacophore Modeling and Virtual Screening Platform

PharmoSB is a direction-aware structure-based pharmacophore modeling and virtual screening platform designed for rapid, reproducible identification of bioactive compounds from large chemical libraries. The software automatically extracts pharmacophoric features from protein–ligand complexes and constructs compact, interaction-focused models suitable for high-throughput screening workflows. PharmoSB integrates vector-aware hydrogen bond feature modeling, steric exclusion volume generation, and conformer-based ligand matching into a unified graphical interface. The platform supports end-to-end workflows including model generation, interactive visualization, virtual screening, and enrichment-based validation using active–decoy datasets. Key capabilities include: Structure-based pharmacophore extraction from PDB complexes Direction-aware HBA/HBD feature modeling Automatic exclusion volume generation for pocket realism Multi-mode virtual screening (Fast, Standard, Deep) Rank-based hit prioritization using pharmacophore fitness Built-in validation via enrichment factor (EF) analysis Publication-quality pharmacophore visualization PharmoSB is implemented in Python with RDKit and provides a user-friendly desktop interface for both novice and advanced users in computer-aided drug design (CADD). The platform is particularly suited for early enrichment studies, hit identification, and academic virtual screening campaigns. This release (v1) establishes the core PharmoSB engine and graphical workflow. Future versions will incorporate consensus pharmacophores, population-weighted scoring, and expanded benchmarking capabilities.