BioDrugScreen

BioDrugScreen ranks docked small molecules from the NCI diversity set against human protein cavities to generate scored receptor–ligand complexes for computational drug discovery and proteome-wide target profiling.

Key Features:

• Compound Library: Nearly 1,600 molecular structures from the NCI diversity set were docked as ligands.
• Targets and Cavities: Docking was performed against 1,589 human protein targets and 1,926 identified cavities.
• Receptor–Ligand Complexes: The dataset comprises over three million receptor–ligand complexes generated by large-scale docking.
• Scoring Functions: Multiple scoring functions were applied, including AutoDock, DockScore, ChemScore, X-Score, GoldScore, DFIRE, and PMF.
• Advanced Rescoring: GBSA and PBSA rescoring were applied for more intensive energetic evaluation of complexes.
• Data Sources: The dataset is derived from an updated Human Cancer Protein Interaction Network and a newly created Human Druggable Proteome.
• Computational Resources: Docking required over 200,000 CPU‑hours and GBSA/PBSA rescoring required an additional 120,000 CPU‑hours on the TeraGrid.

Scientific Applications:

• Virtual Screening and Prioritization: Ranking and prioritization of candidate therapeutic molecules against human protein targets.
• Proteome-wide Target Profiling: Profiling compound binding across 1,589 human protein targets and 1,926 cavities.
• Oncology Research: Analysis of cancer-related targets and compounds using the Human Cancer Protein Interaction Network-derived dataset.

Methodology:

Structure-based docking of ~1,600 NCI diversity set molecules onto 1,926 cavities across 1,589 human protein targets, scoring with AutoDock, DockScore, ChemScore, X-Score, GoldScore, DFIRE, and PMF, with additional GBSA and PBSA rescoring; datasets derived from an updated Human Cancer Protein Interaction Network and a newly created Human Druggable Proteome; computational efforts exceeded 200,000 CPU‑hours for docking and an additional 120,000 CPU‑hours for GBSA/PBSA on the TeraGrid.