dockNmine

dockNmine integrates and analyzes virtual and experimental protein-ligand interaction data to support evaluation of docking results and correlation with experimental binding measurements.

Key Features:

• Data integration and automated querying: Automates queries on protein targets and ligands using Uniprot, PubChem, and ChemBL to enrich pre-computed docking results.
• Identifier standardization and expert annotation: Supports expert annotation and controlled identifiers such as Uniprot IDs for proteins and SMILES for ligands to standardize names and enable cross-database retrieval.
• Private dataset integration: Integrates private experimental datasets with virtual screening and docking results for combined analysis.
• Pre-computed analytical outputs: Provides pre-computed outputs to assess correlations between docking scores and experimental data, including metrics used for ROC curve and enrichment analysis.
• Docking parameter parsing: Parses essential docking parameters automatically from incorporated docking experiments.

Scientific Applications:

• Virtual screening enhancement: Combines virtual screening results with experimental data to improve prediction and evaluation of ligand binding modes on protein targets.
• Data standardization and annotation: Addresses molecule name inconsistencies and transferability issues by standardizing identifiers and annotations across datasets.
• Interaction evaluation and benchmarking: Supports ROC curve generation, enrichment analysis, and correlation assessment to evaluate predictive accuracy of docking in drug discovery workflows.

Methodology:

Queries public databases (Uniprot, PubChem, ChemBL) to gather protein and ligand information, applies expert annotation using identifiers such as Uniprot IDs and SMILES for standardization, and parses essential docking parameters automatically to produce pre-computed outputs for correlation assessment.