Machine Learning-Driven Phenotyping Platform for Rapid In Vivo Target Validation and Precision Medicine

Explore a cutting-edge machine learning-driven phenotyping platform for rapid in vivo target validation and precision medicine in this conference talk from the Toronto Machine Learning Series. Discover how zebrafish are utilized as model organisms in developmental biology, molecular genetics, and toxicology studies due to their unique characteristics. Learn about the challenges in large-scale phenotypic screening for drug discovery and how they are addressed through innovative solutions. Delve into the development of a comprehensive zebrafish phenotyping assay that combines custom 3D-printed mounting platforms with a sophisticated computer vision analysis pipeline. Understand the application of fine-tuned instance segmentation models based on Detectron2 architecture for automatic detection of zebrafish embryos and organs. Examine the platform's effectiveness in studying genes associated with cardiovascular disease using CRISPR injections, and its potential to accelerate drug discovery programs with improved clinical translation prospects.