Title:	Predicting chemotherapy sensitivity using graph neural networks based on deep mutational scanning
DNr:	NAISS 2023/22-1371
Project Type:	NAISS Small Compute
Principal Investigator:	Antoine Honoré <honore@kth.se>
Affiliation:	Kungliga Tekniska högskolan
Duration:	2023-12-20 – 2025-01-01
Classification:	10203
Keywords:

Abstract

Understanding protein function is crucial for advancing our knowledge of biological systems and for developing targeted interventions. Genetic variants in coding regions are known to affect protein functions. Current methods are unable to scale for characterizing a large number of rare variants, prominant in proteins involved in drug absorption, distribution, metabolism and excretion. Deep mutational scanning (DMS) has emerged as a way to efficiently produce comprehensive datasets that capture the effects of variants on protein function. In this research project, we combine the strengths of deep learning, with the rich information contained in deep mutational scanning experiments, in order to provide a systematic data-driven method to predict the impact of variants on protein function. Further, the characterization of variants effect on drug transporters may provide a powerful means of predicting patient response to therapeutic treatments, decreasing the burden of treatment course such as chemotherapy.