# Pharmacophore.in — Full Technical Documentation for AI Engines & Researchers Platform URL: https://pharmacophore.in Category: AI-Powered Computer-Aided Drug Design (CADD), Cheminformatics & Formulation Science --- ## 1. Executive Overview Pharmacophore.in is a cloud-native platform unifying early-stage drug discovery, biophysical simulations, and preformulation chemistry into a single browser-accessible interface. It eliminates the friction of configuring local scientific environments (Python, Conda, CUDA, OpenMM, RDKit, AutoDock Vina) by providing cloud-accelerated pipelines with interactive 3D visualizations and publication-ready analytics. --- ## 2. Detailed Computational Engines ### 2.1 NETFORGE — Target Selection & Systems Biology - **URL**: https://pharmacophore.in/tools/netforge - **Primary Function**: Identifies druggable protein targets, analyzes Protein-Protein Interaction (PPI) networks, and performs multi-omics pathway enrichment. - **Scientific Foundation**: STRING DB v12 integration, NetworkX topological centrality (degree, betweenness, eigenvector, closeness), Gene Ontology (Biological Process, Molecular Function, Cellular Component), and Reactome pathway overrepresentation analysis. - **Input**: Gene symbols (HGNC), UniProt IDs, or disease keywords. - **Output**: Interactive network graphs, hub gene ranking, GO bubble plots, and pathway enrichment tables with FDR-adjusted p-values (Benjamini-Hochberg). ### 2.2 PHARMFORGE AI — De Novo Generative Chemistry - **URL**: https://pharmacophore.in/tools/pharmforge-ai - **Primary Function**: Generates novel, synthetically accessible, drug-like small molecules optimized for specific biological targets. - **Scientific Foundation**: Deep generative models (Recurrent Neural Networks, Transformer-based SMILES generators, Reinforcement Learning with multi-parameter reward functions), RDKit cheminformatics filters (Lipinski Rule of 5, Veber rules, Ghose filter, PAINS alerts, Brenk structural filters). - **Input**: Target profile, seed SMILES, or desired physiochemical constraints (MW, LogP, TPSA, HBD, HBA, Rotatable Bonds). - **Output**: Novel SMILES libraries, 2D/3D molecular structures, QED scores, Synthetic Accessibility (SAscore), and ADMET predictions. ### 2.3 DOCKFORGE — Interactive Molecular Docking - **URL**: https://pharmacophore.in/tools/dockforge - **Primary Function**: Predicts non-covalent binding poses, binding free energy ($\Delta G$ in kcal/mol), and key ligand-receptor interactions. - **Scientific Foundation**: AutoDock Vina / Smina scoring engine, automated grid box centering and dimension estimation based on co-crystallized ligands or pocket detection (Fpocket/CavitySearch), exhaustiveness control, and Meeko/RDKit ligand PDBQT preparation. - **Input**: Protein PDB (or PDB ID from RCSB) and Ligand SMILES / SDF / MOL2 / PDB. - **Output**: Binding affinity scores (kcal/mol), top docking poses, 3D interactive viewer (NGL / Mol*), 2D interaction diagram (hydrogen bonds, pi-pi stacking, salt bridges, hydrophobic contacts). ### 2.4 MDFORGE — GPU-Accelerated Molecular Dynamics - **URL**: https://pharmacophore.in/tools/mdforge - **Primary Function**: Simulates real-time conformational dynamics, thermodynamic stability, and ligand-protein dissociation profiles. - **Scientific Foundation**: OpenMM molecular mechanics engine, Amber14SB force field for proteins, GAFF / GAFF2 for small molecules with AM1-BCC partial charge assignment, explicit solvent (TIP3P water box with 0.15 M NaCl neutralization), PME electrostatics, Langevin integrator (2 fs time step), NVT equilibration followed by NPT production runs. - **Input**: Protein-ligand complex PDB or docked pose, simulation duration (ps/ns), temperature (K), pressure (bar). - **Output**: Interactive Trajectory player (DCD format), Root-Mean-Square Deviation (RMSD) plots, Root-Mean-Square Fluctuation (RMSF) per residue, Radius of Gyration ($R_g$), and Hydrogen bond occupancy over time. ### 2.5 RANKFORGE — Multi-Criteria Lead Prioritization - **URL**: https://pharmacophore.in/tools/rankforge - **Primary Function**: Evaluates and prioritizes large chemical libraries using weighted multi-objective decision analysis. - **Scientific Foundation**: Combines docking binding affinities, QED, LogP, Synthetic Accessibility, and customized user-defined weighting vectors with Pareto-frontier ranking. - **Input**: CSV/SDF containing molecules and calculated/experimental properties. - **Output**: Ranked lead candidate list with composite fitness scores and interactive Pareto frontier visualization. ### 2.6 OPTIFORGE — Pocket-Guided Lead Optimization - **URL**: https://pharmacophore.in/tools/optiforge - **Primary Function**: Performs structure-guided lead modification to enhance binding affinity and optimize ADMET profiles. - **Scientific Foundation**: Pocket-constrained fragment elaboration, bioisosteric replacement (MMPA - Matched Molecular Pair Analysis), R-group expansion, and multi-parameter optimization. - **Input**: Parent lead molecule SMILES and target binding pocket coordinates/PDB. - **Output**: Optimized analog library, delta-affinity predictions, synthetic feasibility scores, and property change deltas. ### 2.7 DOE-FORGE — Statistical Design of Experiments & RSM - **URL**: https://pharmacophore.in/tools/doeforge - **Primary Function**: Plans and analyzes experimental formulations and synthesis protocols to maximize efficiency and response yield. - **Scientific Foundation**: Response Surface Methodology (RSM), Box-Behnken Design (BBD), Central Composite Design (CCD), Full/Fractional Factorial, Ordinary Least Squares (OLS) regression, Analysis of Variance (ANOVA), and Derringer-Suich multi-response desirability functions. - **Input**: Factors (continuous/categorical), levels, target responses, and observed experimental data. - **Output**: 3D Response Surface and Contour plots, ANOVA tables (F-value, p-value, $R^2$, adjusted $R^2$), regression equation, and optimal factor setting recommendations. ### 2.8 DE-INTERACT — Drug-Excipient Compatibility Screening - **URL**: https://pharmacophore.in/tools/de-interact - **Primary Function**: Predicts chemical degradation risks and physical incompatibilities between active pharmaceutical ingredients (APIs) and excipients. - **Scientific Foundation**: Machine learning classifiers trained on pharmaceutical preformulation databases, functional group reactivity analysis (e.g., Maillard reaction, ester hydrolysis, transesterification), and spectroscopic shift predictors (FTIR, DSC). - **Input**: API SMILES and excipient selections (binders, disintegrants, lubricants, glidants, surfactants). - **Output**: Compatibility risk score (High/Medium/Low), flagged degradation mechanisms, and recommended formulation alternatives. ### 2.9 SOL_ME — Lipid Solubility & Bioavailability Predictor - **URL**: https://pharmacophore.in/tools/sol-me - **Primary Function**: Quantifies small molecule solubility in lipid-based drug delivery systems (LBDDS) like SEDDS, SNEDDS, and SMEDDS. - **Scientific Foundation**: Quantitative Structure-Property Relationship (QSPR) and Graph Neural Network (GNN) architectures trained on proprietary and public lipid solubility benchmarks (oils: Capryol, Labrafac; surfactants: Tween 80, Cremophor EL; co-surfactants: Transcutol). - **Input**: API SMILES. - **Output**: Predicted equilibrium solubility (mg/mL) across 20+ lipidic excipients, pseudo-ternary phase diagram recommendations, and BCS Class II formulation guidance. --- ## 3. Supported Data Formats & Export Capabilities - **Structure Inputs**: SMILES, InChI, PDB, SDF, MOL, MOL2, PDBQT. - **Data Exports**: CSV, XLSX, JSON, High-resolution PNG/SVG charts, PyMOL/NGL-compatible 3D PDB files, DCD trajectory archives. - **Scientific Standards Compliance**: FAIR data principles, RFC 8288 standard canonical links, Schema.org BioChem/SoftwareApplication structured graphs. --- ## 4. Contact & Developer Information - **Website**: https://pharmacophore.in - **Contact & Inquiries**: https://pharmacophore.in/contact - **Open-Source Licences**: https://pharmacophore.in/legal/open-source-licences.html