New York: Smita Krishnaswamy, an Indian-origin associate professor of computer science at Yale University, has been named one of the 2025 awardees by Yale’s Colton Center for Autoimmunity. Her research project, ImmunoStruct, is one of nine selected this year for funding support from the Center, which is backing innovative work on autoimmune diseases with nearly $1 million in grants.
Krishnaswamy’s project, developed in collaboration with immunologist Akiko Iwasaki, aims to address a key challenge in immunotherapy and vaccine development: predicting which parts of a virus or cancer cell the immune system will recognize and respond to—known as epitopes. Her team’s solution, ImmunoStruct, is a deep-learning model that goes beyond existing tools by incorporating not just the amino acid sequence of proteins, but also their 3D structure and chemical properties.
What this really means is: instead of relying only on the genetic code, ImmunoStruct uses a more complete picture—how these protein fragments are shaped and behave chemically in the body. The model was trained on about 27,000 known peptide-MHC (major histocompatibility complex) combinations and has already shown stronger predictive accuracy than current tools. It aligns closely with lab test results for COVID-19 epitopes and has also shown potential in identifying cancer-related immune responses.
