05/01/2019 12:07 AM IST | Updated 05/01/2019 12:07 AM IST

Food Scientists Say AI Can Give Ayurveda Scientific Rigour

Geneticists and chemical engineers believe that Ayurveda contains observational science that remains valuable—but only if it is thoroughly interrogated using the techniques of modern medicine.


NEW DELHI —“Your body is a complex microbiome, adapted by the food you eat, and the environment you’re in,” said Mitali Mukerji, Senior Principal Scientist at the CSIR Institute of Genomics and Integrative Biology.

Mukerji, who is best known for coining the term ayurgenomics—blending Ayurveda with modern genomics—explained that the food we eat every day can have a significant impact on our bodies and our minds, beyond the principles of nutrition.

“The health effects of different types of foods are well known and well studied, but what’s less understood, in many cases, is why this is so,” Mukerji said.

Ayurvedic texts might suggest eating foods like turmeric to bolster the immune system, and this is something that has now caught on around the world, but Mukerji is applying the scientific method to break down how exactly this works so that people can make informed decisions about their health. By bringing together Ayurveda and scientific rigour, the hope is to develop cures that can be tailored for an individual, while avoiding negative effects or ineffective dosages.


Mukerji spoke at the sidelines of the Symposium on Computational Gastronomy at IIIT-Delhi. At the symposium, Ranjan Bose, Director IIIT-Delhi explained that we are living at a time where the availability of data is unprecedented, and what’s required is to find ways to analyse information and derive a better understanding of complex systems like food.

This extends not just to information such as the chemical and biological composition of food and its effects on the body, but also data such as the individual preferences and tastes of people.

“Today, food is digitised. Our tastes, our preferences, our consumption is digitised. People can’t eat without first taking pictures for social media. Restaurant listings have detailed menus, and you can also see reviews and feedback, it is all on the Internet. And we can mine this data, and crowdsource data, and analyse it to transform food.”

The use of AI and machine learning, Bose said, can enable researchers to finally make sense of this vast amount of data.

Applying data

One of the ways in which researchers at IIIT-Delhi are trying to do this is by using AI and machine learning to find new combinations of flavours. But another development from the research lab led by Ganesh Bagler at IIIT-Delhi is called DietRx, a database that collects information about the molecules present in different ingredients, and links that to medical reactions, chemicals in medicines, and genetic reactions.

If you visit the DirectRx website created by Bagler and his research students, you’ll be able to look up specific ingredients—let’s say ginger—and you can see different diseases and the associations that the chemicals inside ginger have, both positive and negative. So, skimming through the list, you can see that it helps with nausea, vomiting, inflammation, and pain. On the other hand, if you open up nausea, you’ll see that cardamom and piper nigrum (black pepper) are listed as having negative associations. This kind of information is useful because if you’re making food for someone suffering from stomach cramps, for example, you don’t want to give them something that makes matter worse. This could be very useful also for caregivers to people with chronic diseases.


These are all things that become possible only through the rigorous scientific questioning of the effects of different ingredients. “We have this long history of knowledge, which is amazing, but we need to be asking questions and studying it carefully to bring it to the global stage now,” said Mukerji.

As Bagler points out, people want to know answers to questions like “is red wine good for you? Should you have more coffee, or less coffee? What about eggs?” These are questions that have importance not just for nutrition, but in a country like India, also hold political significance.

“Food is also very complex, so different combinations of ingredients can have different effects as well. This is a very complex problem. We are like the blind men and the elephant. We need mathematical and biological models, based on exhaustive data collection, and AI techniques to analyse it.”

Deciphering the ‘RGB codes’ of food

That’s where people like Mukerji come in. By studying the impact of Ayurvedic medicines using genetics research, his team hopes to break down the interactions taking place in a scientific manner. Rather than relying on observations made centuries ago, and anecdotal learning, the research aims to show exactly what mechanisms are being activated, and how different foods are affecting the body.


“It’s like with a computer, everything is just code, or colors are RGB codes, right? So that’s what we are doing now, we’re building the RGB codes in the language of molecular biology, from the original Sanskrit,” said Mukerji. By collecting traditional knowledge and using AI to compare it against chemical and molecular reactions that have been studied, and tracking vast amounts of complex genetic data, scientists can bring the rigor that is needed.

Doing this could also have an impact on not just medical research, but also everyday food items. If scientists can find a direct, provable link between certain food items and health, beyond nutritional information like salt, fat, sodium, and so forth, then the FMCG companies that make food items can also take advantage of the information to create healthier alternatives, suggested Gurmeet Singh, a chemical engineer with over 20 years of experience at Unilever, where he worked as global research and development director. He is a professor and head of center for ayurveda, biology, and holistic nutrition consultant to tea, beverages and foods industry at the Trans Disciplinary University in Bengaluru.


Singh agreed with Mukerji that there is a lot of value in researching the health impacts of different types of foods, and explained how even among largely understood foods like buttermilk, ratios of ingredients, stages of fermentation, and even temperature at the time of consumption can lead to different impacts on the body. “All of this needs to be scientifically researched, and the food industry can do more.”

“Today the guidelines of the food advice we receive are too broad, and as a result, the body is under stress,” he said. “And when it comes to the packaged food industry, when food is being designed, the priorities are cost, then sensory pleasure, then nutrition, then finally you might think of health. We need to turn that around completely.”