Systems Pharmacology-Informed Phytotherapy Analysis of Triphala: A Complete Computational Dataset

Note: This dataset has been revised in response to peer review. It now includes the updated CABS-flex 2.0 methodology, the complete composite scoring methodology (Supplementary Methods S1), the explicit ADMET compound filtering criteria (Supplementary Table S10), and the variant-stratified compound prioritization table (Supplementary Table S11). All file structures and validation metrics have been synchronized with the revised manuscript. This dataset provides the complete computational pipeline, raw data, statistical validations, and source code accompanies the manuscript entitled: "Systems Phytotherapy for Neuro-Immune Crosstalk: Triphala Phytochemicals as Predicted Modulators of Inflammation-Driven Neurodegeneration via the Melatonin Pathway" (Tentative; Manuscript ID: 1892413; Journal: Frontiers in Drug Discovery; Topic: Anti-inflammatory and Immunomodulating Agents) investigating the neuro-immune and anti-inflammatory potential of Triphala using systems pharmacology and molecular dynamics. To ensure complete transparency, reproducibility, and alignment with open science principles, this repository contains every intermediate and finalized data structure utilized in the study, ranging from raw phytocompound SMILES to coarse-grained MD simulation trajectories. Dataset Structure The archive is comprehensively structured into 12 distinct directories: 01_Compound_Library_and_SMILES: Curated library of phytocompounds identified in Triphala extracts, including canonical SMILES required for predictive modeling. 02_Target_Prediction_and_ADMET_Outputs: Raw probability scores from SwissTargetPrediction and comprehensive ADMET evaluations from pkCSM, SwissADME, and AdmetSAR. 03_Pathway_Enrichment_Tables: Complete Gene Ontology (GO) and KEGG pathway enrichment analysis tables, including FDR-corrected p-values. 04_Molecular_Docking_Data: Complete AutoDock Vina datasets, including receptor structures (.pdbqt), grid parameter files (conf.txt), scoring logs, and the best docked ligand poses (.sdf and .pdbqt). 05_MD_Simulation_RMSF_Files: Root Mean Square Fluctuation (RMSF) trajectories from CABS-flex 2.0, mapping the structural dynamics of the top receptor-ligand complexes. 06_Analysis_Scripts: Custom Python scripts used to compute the Composite Pathogenicity Index (CPI), calculate Ligand Efficiency, perform network topology ranking, and render publication-quality figures. 07_Supplementary_Figures: High-resolution supplementary figures (S1–S32), including statistical benchmarking graphs and docking distributions. 08_Statistical_Validation: Statistical reports validating docking scores, ligand efficiency, and structure-function relationships. 09_Network_Files_Supplementary_Data_S3: Cytoscape session files (.cys) and node/edge attribute tables used for centrality mapping of the neuro-immune interactome. 10_Supplementary_Tables_and_Reports: Formatted tables and methods documents (S1–S11), including variant allele frequencies, ADMET filtering criteria, hub protein centralities, and re-docking RMSD validation reports. 11_wwPDB_Validation_Reports: Official wwPDB validation reports assessing the structural quality of experimental protein models used in docking. 12_Supplementary_Data: The consolidated primary data packages (Data S1–S5) provided during peer review. Additional Files Total Timeline of the Project including Variant Curation Protocol.md: Detailed timeline capturing database access dates (e.g., ClinVar, gnomAD, KEGG, PubChem) utilized during variant curation. README_Zenodo_Supplementary_Data.md: A guide to navigating the directory structures and understanding data provenance. Software and Tools Used The following software and web-based computational tools were utilized to generate this dataset: AutoDock Vina & MGLTools: For molecular docking simulations and preparation of .pdbqt files. CABS-flex 2.0: For coarse-grained molecular dynamics simulations and RMSF trajectory calculations. Cytoscape 3.10: For construction and topological analysis (degree/betweenness centrality) of the neuro-immune target networks. Python 3.x: For statistical validations, automated data parsing, Ligand Efficiency calculations, and rendering high-resolution plots (using matplotlib, seaborn, pandas). SwissTargetPrediction, SwissADME, pkCSM, & AdmetSAR: Web servers used for pharmacological target predictions and ADMET property evaluations. PyMOL / Discovery Studio Visualizer: Utilized for 3D coordinate processing and visualization of receptor-ligand complexes. Reproducibility All Python scripts provided in 06_Analysis_Scripts are fully commented and can be executed to reproduce the statistical figures and network topology calculations presented in the manuscript.