Enhancing the Drug Discovery Process: Using Bayesian Inference for the Analysis of Dose-Response Experiments

Caroline Labelle

When developing a new therapy, many chemical compounds are tested experimentally. Efficient and powerful compounds are then selected fur further exploration. Caroline’s work focus on using a Bayesian framework to infer the effectiveness of compounds based on experimental data. Using probabilistic programming, she proposes a statistical approach to select compounds considering both experimental and analytical noise. She developed a web interface called BiDRA that is simple to use. Preliminary quantitative results suggests that BiDRA is more robust that the standard accepted approach. Caroline’s thesis aims at aims to better equip biologists and medicinal chemists in their decision making process when studying cellular responses to multiple chemical compounds.

Keywords: bayesian inference, probabilistic programming, dose-response, high-throughput screen, compounds selection, drug discovery process, development anlaysis Pipelines, computer sciences, bioinformatics

Sébastien Lemieux
Sébastien Lemieux
Principal Investigator