1 week ago

AI Helps Target ‘Undruggable’ Protein in Pancreatic Cancer

AI Helps Target ‘Undruggable’ Protein in Pancreatic Cancer
Treatment for pancreatic cancer? How AI helped target ‘undruggable’ protein · indianexpress.com

Scientists are studying a protein called GIPC1 that can help some cancer cells grow and spread.

It is found in high amounts in pancreatic ductal adenocarcinoma, an aggressive form of pancreatic cancer.

GIPC1 has been difficult to target because it connects with many other proteins.

Mayo Clinic and Sravathi AI Technology used artificial intelligence to search for a molecule that could block it.

They examined about 40,000 possible molecules and eventually chose one called GIPCi.

Tests so far have included laboratory research and animal trials.

The study suggests that GIPCi may slow tumour growth and improve chemotherapy response in research models.

However, the findings are still preclinical and do not show that the treatment is safe or effective in people.

Researchers estimate human use could be about three years away if further studies are successful.

Key facts

Target protein
GIPC1, a signalling protein associated with cancer growth in certain circumstances.
Candidate molecule
GIPCi, a selective small-molecule inhibitor targeting GIPC1’s PDZ domain.
Cancer focus
Pancreatic ductal adenocarcinoma, or PDAC, the most common and aggressive type of exocrine pancreatic cancer.
Reported survival rate
The article states that PDAC has a five-year survival rate under 13.3%.
AI-assisted screening
Researchers started with about 40,000 molecules and narrowed them to five candidates.
Research status
The findings are preclinical, and Mayo Clinic has continued research including animal trials.
Potential timeline
According to Sravathi AI founder Gurram Kishan, human use could be possible in about three years if the treatment proves suitable.

Quotes

Debabrata Mukhopadhyay

Senior author of the study and cancer researcher at Mayo Clinic Florida

“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.”
indianexpress.com
“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.”
indianexpress.com

Sources

Related news