Data to Therapy (D2T): Revolutionizing the discovery of Targeted Therapy with Artificial Intelligence

Award: Google PhD Fellowship - Mahdieh Labani

Abstract:Target discovery is a process that involves identifying biomarkers in a disease's biological process that can be targeted by a drug. This process is often slow and can take many years to complete. Even once a target has been chosen, identifying molecules with the desired properties can be challenging and uncertain. However, AI-enabled techniques can help overcome these limitations by training on large datasets to identify patterns and trends that may not be visible to humans. This project aims to develop an AI-enabled software that will improve the target discovery process, helping companies identify promising leads and accelerating drug discovery. The research plan involves two phases to detect novel targets that regulate cancer genes using AI. The first phase involves identifying the causal mutations and underlying biological mechanisms and understanding the impact of these variants. Gene expression for individuals with these variants is compared against healthy individuals to validate them as a biomarker. For the complex structure of CNV, the causative CNV is detected in the first objective, followed by the combination of other variants in the second objective to determine their function. In some cases where the gene expression value is missing, the third objective will involve using AI to predict the gene expression value and validate the identified variants as a biomarker.