Leveraging automation and AI for high-throughput discovery of condensate-modifying drugs

Abstract: Biomolecular condensates are reversible, membrane-less organelles with unique compositions that regulate diverse cellular processes including RNA metabolism, transcription, and signal transduction by selectively phase separating and concentrating biomolecules within specific regions of cells. Misregulation of condensates (we call this a condensatopathy) has been implicated in many diseases such as cancer, neurodegeneration, metabolic diseases and infectious diseases. Increasing evidence supports a model where condensates serve as central nodes of dysregulation in complex diseases. Furthermore, condensate communities are enriched in targets classically considered undruggable. Focusing drug discovery efforts on condensates increases the available target space by ~10-fold, while addressing complex sources of dysfunction, and therefore presents novel, untapped opportunities for therapeutic development.

We will present how Dewpoint Therapeutics approaches condensate-targeted drug discovery and discuss initial results from a condensate-based screen. Our fully integrated end-to-end high-throughput screening process is coupled with generative AI to accelerate identification and optimization of condensate-modifying drug (c-mod) candidates. This platform opens new opportunities to discover drugs for complex diseases that were previously thought undruggable.