Abstract
(This speech will be given together with Ms. Sofia Ormazabal Arriagda.)
How can we better predict cancer progression by tapping into the intricate web of gene interactions and the wealth of information within protein sequences? This talk explores an approach that combines Large Language Model embeddings with Graph Neural Networks to capture the complexities of cancer biology. By integrating gene-gene interactions and protein structure data, we can capture patterns and abnormalities present in various cancers, providing a more comprehensive view of a patient's health status and improving their treatment planning. Our model shows improved precision in stratifying patient groups and identifying high-risk individuals, particularly for breast, lung, and colon cancers.
How can we better predict cancer progression by tapping into the intricate web of gene interactions and the wealth of information within protein sequences? This talk explores an approach that combines Large Language Model embeddings with Graph Neural Networks to capture the complexities of cancer biology. By integrating gene-gene interactions and protein structure data, we can capture patterns and abnormalities present in various cancers, providing a more comprehensive view of a patient's health status and improving their treatment planning. Our model shows improved precision in stratifying patient groups and identifying high-risk individuals, particularly for breast, lung, and colon cancers.