Development and application of a novel network pharmacology and reverse docking method for natural product discovery: a case study with Gastrodia elata (Tian Ma)

Abstract

Gastrodia elata, commonly referred to as 'Tian Ma' in China, is a prominent Traditional Chinese Medicine (TCM) utilized in clinical treatments for neurological disorders, hypertension, and other ailments. Despite extensive research on this TCM, numerous studies have primarily focused on its traditional or well-established ingredients, diseases, and pathways. Leveraging advancements in computational techniques, our research endeavors to uncover potential 'Molecule-protein pathway-disease' axes, aiming to bridge the gap between traditional knowledge and modern drug discovery. Our methodology commences with the integration of TCM compound databases and literature reviews to predict Tian Ma's potential efficacious molecules. With our curated list, we employ AlphaFold2 for Reverse Docking—an AI-assisted visual screening—to pinpoint potential target proteins for each molecule. Subsequent integrative analyses involve GO and KEGG functional annotations, complemented by Disease Association Analysis using DisGeNET and Human Phenotype Ontology (HPO) databases. Utilizing Cytoscape software, we visualize the intricate network, striving to identify the potential 'Molecule-protein-pathway-disease' axes. Upon identification of promising axes, we engage in a preliminary 'dry lab' verification using Molecular Dynamics via Gromacs to ascertain the stability of the compounds. Positive outcomes in this phase will guide our evaluation on the viability of advancing to 'wet lab' pharmacological verification, ensuring a comprehensive and methodical approach to harnessing the therapeutic potential of 'Tian Ma'. Furthermore, we also developed a python-based software for quick literature review with NLP and machine learning applied