Exploring AI methods to improve phenotypic drug discovery- Thiago C. Cavalcanti, PhD(AI-CRED Fellow)

Dr. Thiago C. Cavalcanti: https://sites.google.com/view/aicred/... English Summary This presentation introduces the challenges and limitations of modern drug discovery and the potential of phenotypic drug discovery to address them. It discusses how high-content imaging enables detailed characterization of cellular phenotypes while generating large volumes of complex imaging data. The presentation highlights the limitations of conventional image-analysis pipelines and the potential of end-to-end deep learning approaches. Recent methods, including PhenoProfiler and CellCLIP, are presented to demonstrate how deep learning and multimodal learning can improve phenotypic profiling. Finally, the presentation discusses the availability of large-scale bioimage datasets that are enabling further development of AI-based approaches for drug discovery.