Natural products have been a cornerstone in drug discovery, contributing to over 1000 therapeutics in the past three decades. Traditional methods, such as the "grind and find" approach, have evolved with technological advancements, enabling high-throughput and high-content screening of chemical compounds. These advanced screening methods allow for measuring complex phenotypes, particularly through image-based assays that capture subtle cellular morphological changes. However, the complexity of the resulting data necessitates robust algorithms and workflows to profile and interpret these phenotypes accurately. First, I will present my work on refactorizing the HistDiff algorithm, routinely used for the Cytological Profiling assay (a high-content image-based phenotypic screening assay) to define the phenotypic fingerprints of compound treatments. Next, I introduce MOAST (Mechanism of Action Similarity Tool), a BLAST-inspired methodology to associate mechanism of action annotations of known reference compounds to unknowns through the use of pairwise associations of the phenotypic fingerprints and its transition to a formal machine learning classifier framework. Finally, I will outline the design behind a web application that serves as an all-in-one ecosystem for integrating the process of submitting compounds to the UCSC screening center with visualization of the raw phenotypic screening data and annotation of the processed data.

Event Host: Akshar Lohith, Ph.D. Candidate, Biomolecular Engineering & Bioinformatics

Advsiors: Scott Lokey and John MacMillan