The Ministry of Science and Technology said on Wednesday that researchers at the S N Bose National Centre for Basic Sciences presented an artificial intelligence (AI) framework that could revolutionize our knowledge of and approach to treating cancer, as well as open the door to customized treatments.

The framework, which was created in partnership with Ashoka University, offers a fresh perspective on cancer by focusing on its molecular personality rather than just its size or spread.

“The hallmarks of cancer are a collection of hidden biological programs that drive cancer, which is more than just a disease of expanding tumors. According to the Ministry, “these characteristics explain how healthy cells become malignant: how they spread, evade the immune system, and resist treatment.”

Although staging systems like TNM, which characterize the size and distribution of tumors, have been used by doctors for decades, they sometimes overlook the underlying molecular story. Why, for instance, can two people with the “same” cancer stage have such disparate results?

According to the Ministry, a new AI framework called OncoMark can read the molecular “mind” of cancer and forecast its behavior.

Under the direction of Drs. Shubhasis Haldar and Debayan Gupta, the SN Bose team analyzed 3.1 million single cells from 14 different cancer types using OncoMark.

The Department of Science and Technology (DST) has an independent institute called S N Bose.

In order to simulate hallmark-driven tumor states, the study team produced artificial “pseudo-biopsies.”

The AI was able to learn how characteristics like immune evasion, metastasis, and genomic instability interact to promote tumor growth and treatment resistance thanks to this massive dataset.

“OncoMark maintained accuracy above 96% in five separate cohorts and attained above 99% accuracy in internal testing. It demonstrated wide applicability when validated on 20,000 real-world patient samples from eight significant datasets. Scientists were able to see for the first time how signature activity increases as cancer progresses, according to the Ministry.

The new approach identifies the hallmarks that are active in a patient’s tumor and was published in the Nature journal Communications Biology. This can point physicians in the direction of medications that specifically target certain mechanisms.

Additionally, it can promote earlier action by identifying aggressive malignancies that may appear less dangerous under traditional staging, according to the Ministry.