Abstract 4496: Small molecule inhibition of previously “undrubbable” transcription factors with AI-guided functional proteomics

Abstract Transcription factors (TFs) are central regulators of gene expression and key drivers of diseases such as cancer, autoimmune disorders, neurological disease, and metabolic conditions. Despite their critical role, TFs have long been considered “undruggable” due to their intrinsic disorder outside of native cellular environments, complex multiprotein assemblies, and the absence of traditional binding pockets suitable for small molecule targeting. To address these challenges, Talus Bio has developed TF-Scan, a live-cell platform that provides a proteome-wide view of drug-induced changes in TF DNA-binding activity. This technology integrates high-resolution functional proteomics with advanced machine learning to measure TF:DNA engagement directly in unmodified human cells. Data from TF-Scan feeds into Strategian, a deep tensor factorization model trained on Talus Bio’s proprietary TF-Scan database, enabling AI-driven discovery of novel small molecule modulators. Strategian evaluates a chemical universe of billions of drug-like molecules to predict their effects on TF activity, prioritizing candidates by potency and selectivity. These predictions are validated experimentally using TF-Scan, which further refines lead selection. This approach has enabled the identification of novel scaffolds for TFs including TBXT, ARv7, and STAT3, demonstrating the ability to generate drug-like compounds tailored to specific TFs. By combining live-cell proteomics with AI-driven modeling, TF-Scan and Strategian create a powerful, scalable platform for targeting TFs and other challenging regulators within the regulome. This work establishes a new paradigm for drug discovery, transforming previously intractable targets into actionable opportunities to address critical unmet medical needs. Citation Format: Alexander J. Federation, Daniele Canzani, Andrea Gutierrez, Julia Robbins, Brian McEllin, Anastasiya Prymolenna, Lillian Tatka, Bodhi Hueffmeier, Yang Gao, Tonibelle Gatbonton-Schwager, Erin Broderick, Gaelle Mercenne, William Fondrie, Lindsay Pino. Small molecule inhibition of previously “undrubbable” transcription factors with AI-guided functional proteomics [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2025; Part 1 (Regular Abstracts); 2025 Apr 25-30; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2025;85(8_Suppl_1):Abstract nr 4496.