MC EXCLUSIVE AI’s next biological leap could be cracking the genome, says Google DeepMind scientist

• AI seeks to decode the human genome for disease insights.
• AlphaFold has made protein structure prediction faster and more accessible
• AI is democratizing science, giving global researchers access to advanced tools

Did our AI summary help?

AI’s next big biological frontier is not just mapping proteins but understanding the deeper “recipe book of life” hidden inside human DNA, Google DeepMind’s Vice President of Research Pushmeet Kohli said in an interview with Moneycontrol's Managing Editor, Dr Nalin Mehta.

Kohli, whose teams helped build the Nobel Prize-winning AlphaFold model that predicts protein structures, said the next grand scientific challenge for artificial intelligence is deciphering the human genome and interpreting what genetic codes actually mean for diseases, development and treatment.

“If proteins are the building blocks and ingredients of life, the key question now is to decipher the recipe book of life, our genome,” Kohli said. He explained that while scientists have successfully developed technologies to read DNA sequences over the past few decades, understanding their full implications remains incomplete.

According to Kohli, the genome holds crucial information about how humans develop, which diseases they may be susceptible to, and how medical interventions could be tailored. “We are yet to really decipher it and understand the meaning of the genome,” he noted, describing it as one of the biggest scientific challenges currently underway in AI research.

He added that the long-term goal is to understand the effect of every possible change in the genetic code and use that knowledge not only for diagnosing diseases but also for developing more precise treatments.

Kohli also said AI’s growing role in biology is already demonstrating how technology can level the global scientific playing field. Citing AlphaFold as an example, he noted that determining the structure of a single protein earlier took years and millions of dollars in laboratory work. Today, researchers can obtain similar structural insights in seconds through AI databases.

This shift, he said, is helping democratise science by giving smaller research institutions and scientists in developing countries access to tools that were once limited to well-funded laboratories. “All of that millions of dollars and multiple years has been transformed into one click,” Kohli said, adding that such acceleration shows how AI can expand scientific opportunity across borders.