Title:Design of Novel Drug-like Molecules Using Informatics Rich Secondary Metabolites Analysis of Indian Medicinal and Aromatic Plants
Volume: 23
Issue: 10
Author(s): Divya Karade, Durairaj Vijayasarathi, Narendra Kadoo, Renu Vyas, P.K. Ingle and Muthukumarasamy Karthikeyan*
Affiliation:
- Chemical Engineering and Process Development (CEPD) Division, CSIR-National Chemical Laboratory, Pune - 411008,India
Keywords:
Medicinal plants, metabolites, text mining, drugs, scaffolds, virtual libraries, virtual screening.
Abstract:
Background: Several medicinal plants are being used in Indian medicine systems from ancient
times. However, in most cases, the specific molecules or the active ingredients responsible for the medicinal
or therapeutic properties are not yet known.
Objective: This study aimed to report a computational protocol as well as a tool for generating novel
potential drug candidates from the bioactive molecules of Indian medicinal and aromatic plants through the
chemoinformatics approach.
Methods: We built a database of the Indian medicinal and aromatic plants coupled with associated
information (plant families, plant parts used for the medicinal purpose, structural information, therapeutic
properties, etc.) We also developed a Java-based chemoinformatics open-source tool called DoMINE
(Database of Medicinally Important Natural products from plantaE) for the generation of virtual library and
screening of novel molecules from known medicinal plant molecules. We employed chemoinformatics
approaches to in-silico screened metabolites from 104 Indian medicinal and aromatic plants and designed
novel drug-like bioactive molecules. For this purpose, 1665 ring containing molecules were identified by
text mining of literature related to the medicinal plant species, which were later used to extract 209
molecular scaffolds. Different scaffolds were further used to build a focused virtual library. Virtual screening
was performed with cluster analysis to predict drug-like and lead-like molecules from these plant molecules
in the context of drug discovery. The predicted drug-like and lead-like molecules were evaluated using
chemoinformatics approaches and statistical parameters, and only the most significant molecules were
proposed as the candidate molecules to develop new drugs.
Results and Conclusion: The supra network of molecules and scaffolds identifies the relationship between
the plant molecules and drugs. Cluster analysis of virtual library molecules showed that novel molecules had
more pharmacophoric properties than toxicophoric and chemophoric properties. We also developed the
DoMINE toolkit for the advancement of natural product-based drug discovery through chemoinformatics
approaches. This study will be useful in developing new drug molecules from the known medicinal plant
molecules. Hence, this work will encourage experimental organic chemists to synthesize these molecules
based on the predicted values. These synthesized molecules need to be subjected to biological screening to
identify potential molecules for drug discovery research.