Pranshi Mehta - Society for Science Skip to content

Pranshi Mehta

7th Grade, Canyon Vista Middle School
Austin, TX

Machine Learning-Mediated Computational Modeling of FK506-Binding Protein 12 (FKBP12)-Enhanced CAR T-Cell Therapy for Targeted Glioblastoma Treatment

View Poster
2025 Thermo Fisher JIC Pranshi Mehta: Machine Learning-Mediated Computational Modeling of FK506-Binding Protein 12 (FKBP12)-Enhanced CAR T-Cell Therapy for Targeted Glioblastoma Treatment
Machine Learning-Mediated Computational Modeling of FK506-Binding Protein 12 (FKBP12)-Enhanced CAR T-Cell Therapy for Targeted Glioblastoma Treatment Pranshi Mehta
View Poster

Project Background

When Pranshi’s best friend lost her mother to cancer, Pranshi began to wonder why some cancers can be treated successfully while others can’t. She discovered CAR T-cell therapy—which takes the immune system T cells from a cancer patient and genetically alters them to produce antibodies to attack the cancer. CAR T-cell therapy works well for blood cancers, but not in solid cancers like glioblastoma. “I wanted to understand why this therapy worked in some cancers but not others, and whether artificial intelligence tools could help find answers,” Pranshi says.

Tactics and Results

Pranshi decided to focus on glioblastoma, one of the most aggressive and lethal brain cancers. She wanted to look for potential antibodies that could bind to FKBP12, a protein that is overexpressed in glioblastoma. She then used the modeling program AlphaFold3 to generate a 3D structure of FKBP12, to select the best regions for an antibody to bind. Pranshi picked 28 antibodies that might fit and used a protein simulator to see how well each one would fit on to the protein and calculated how strong the binding would be. She ended up with one antibody called 3D85, which stayed bonded to FKBP12 for 10 nanoseconds. Her result shows a way to use AI to find new antibodies quickly. The next step, Pranshi says. “is to experimentally validate these antibody candidates and integrate them into functional CAR constructs.”

Pranshi Mehta
Lisa Fryklund/Licensed by Society for Science

Beyond the Project

Pranshi plays the piano, performs monthly at retirement homes, and has performed twice at Carnegie Hall. She loves to write and is currently working on her third book. Pranshi also likes to go on bike rides with her friends, and to swim. “I swim to clear my head, and the rhythm of the laps has sparked many of my best Eureka moments, including my research,” she says. She would like to go into health care. “Health care combines science, art, and service, my three passion points,” she says.

2025 Thermo Fisher JIC Finalist Pranshi Mehta
Pranshi Mehta