Renowned US-based medical centre Mayo Clinic recently announced that it has co-developed a cancer treatment with Bengaluru-based artificial intelligence (AI) startup Sravathi AI Technology, under an intellectual property sharing arrangement. This treatment targets a specific protein called GIPC1 that, in certain circumstances, can contribute to the growth of many different types of cancer, including pancreatic cancer.

The findings were published in the paper, “AI-driven discovery and validation of a GIPC1 PDZ domain inhibitor for pancreatic ductal adenocarcinoma”, in the scientific journal Cell Reports.

According to Sravathi’s founder Gurram Kishan, the GIPC1 protein was considered a mostly “undruggable” target, and that Mayo Clinic had been working on the problem for 12 years when they started collaborating with them.

Story continues below this ad

Here’s what to know about the GIPC1 protein, its links to cancer, and how the treatment targets this protein.

Why GIPC1 is difficult to target

GIPC1 does not cause cancer on its own. While its presence in the body is not abnormal, it is associated with several different types of cancer in certain circumstances. Experts from Sravathi AI noted that when there is a problem in the expression or signalling of this protein in the body, then cancer cells can use the “pathways” associated with this protein to grow excessively or invade other parts of the body.

Pancreatic ductal adenocarcinoma (PDAC) — the most common and aggressive type of exocrine pancreatic cancer — remains highly aggressive, with a five-year survival rate under 13.3% owing to late diagnosis, rapid progression, and therapeutic resistance. GIPC1, which is overproduced in PDAC, drives tumour growth and resistance to chemotherapy but has remained “undruggable” due to its PDZ domain.

The PDZ domain is made up of amino acids (also the building blocks of proteins) and is found in the signalling proteins of bacteria, yeast, plants, viruses, and animals.

Story continues below this ad

On why this protein is so hard to target during treatment, Sravathi CEO Parag Tipnis said: “Its PDZ domain interacts with multiple proteins through broad, shallow surfaces, making it difficult for conventional small molecules to bind effectively. It is also a signalling hub, so blocking one pathway may not be enough to stop the others that remain active.”

The treatment

Using advanced computational modeling, machine learning, and predictive analytics, a selective and specific small molecule inhibitor of GIPC1 was identified, targeting its PDZ domain. This molecule inhibitor is called GIPCi.

Kishan said, “We started with around 40,000 molecules and progressively narrowed them down to five candidates, of which two were synthesised and tested. Generative AI helped us design the initial molecules, while predictive AI was used to assess properties such as toxicity and absorption.”

“We then used molecular modelling and quantum chemistry to identify the molecules most likely to bind GIPC1, ultimately leading us to the final candidate,” Kishan added.

Timeline for human use

Story continues below this ad

Sravathi AI began work on discovering the molecule in 2023, completing their research in about five months. The Mayo Clinic then continued its research, which included animal trials. According to Kishan, if the treatment turns out to be suitable for human use, it could be in use after about three years.

Debabrata Mukhopadhyay, senior author of the study and a cancer researcher at the Mayo Clinic Florida, said in a statement: “Our study demonstrates that AI can help us identify entirely new therapeutic opportunities against targets that have historically been considered undruggable. While these findings are preclinical, they provide a strong foundation for the next phase of research.”