Globally, mankind faces the challenge of “diabesity” — the dual problem of type 2 diabetes and obesity. In 2024 alone, diabetes claimed an estimated 3.4 million lives globally, signalling the limitations of current treatments.

The pills in use typically address one symptom or molecular pathway. Now, a computational study by the Institute of Mathematical Sciences (IMSc) and IIT-Madras suggests that ayurvedic medicine offers a ‘networked’ architecture for treatment.

Led by Prof. Areejit Samal, the research team explains at a molecular-level why herbs used in ayurvedic treatments are effective. “We used advanced knowledge graphs and network pharmacology,” Samal says. .

The study used the IMPPAT database built earlier by Samal’s group.

Drug cocktail

The study focused on the concept of ‘single herbal drugs’ (SHDs). Samal explains that, unlike what the name may suggest, these herbs are actually phytochemical cocktails. Unlike a purified pharmaceutical molecule, a herb such as Aegle marmelos (bael) or Terminalia arjuna contains an array of plant chemicals — ‘phytochemical cocktails’ — that work in synergy.

The researchers studied 11 SHDs indicated in the Ayurvedic Pharmacopoeia of India for both diabetes and obesity. This set of drugs contains 188 bioactive phytochemicals.

‘Knowledge graph’

The team used ‘knowledge graph’ analyses to map the chemicals to 1,099 unique human protein targets, generating information on close to 4,000 biological interactions.

A ‘knowledge graph’ essentially organises data from different sources in a manner that the links between different entities. become evident. The entities are called ‘nodes’ and the relationships are referred to as ‘edges’. In Samal’s study, the drugs were the nodes and the therapeutic pathways were the edges.

The study found that the herbs work as ‘natural polypills’ — each herb contains chemical pairs that attack various protein targets, addressing the same therapeutic pathways such as insulin signalling and lipid metabolism.

This ‘multi-target’ mechanism is similar to combination therapies in modern medicine, such as the metformin–sitagliptin or semaglutide–metformin combinations. Samal says the SHDs may have effects similar to FDA-approved combination drugs, which help address chronic metabolic disorders more effectively than a single-target approach.

When looking at disease targets such as GLP1R — the receptor targeted by weight-loss drugs — the study found that eight of the 11 herbal formulations had at least one phytochemical known to bind to this receptor.

Lead molecules

The team conducted digital molecular simulations to measure how well the natural compounds bind to human proteins

However, isolating and using a purified extract of a specific compound found effective in ayurveda may not be ideal for drug formulations. Samal says the safety and efficacy of ayurveda lie in the complex interaction of whole formulations, rather than isolated ingredients, such as curcumin in turmeric. Turmeric is good and a part of Indian cuisine, but purified curcumin may present safety issues.

Asked if the knowledge graph could be used for the study of drugs designed for other conditions such as liver diseases, Samal says the knowledge graph sets the template for developing other graphs for specific conditions.

More Like This

Published on March 9, 2026