Enhancing the\nPredictive Power of Machine Learning\nModels through a Chemical Space Complementary DEL Screening Strategy

DNA-encoded library (DEL) technology\nis an effective\nmethod for\nsmall molecule drug discovery, enabling high-throughput screening\nagainst target proteins. While DEL screening produces extensive data,\nit can reveal complex patterns not easily recognized by human analysis.\nLead compounds from DEL screens often have higher molecular weights,\nposing challenges for drug development. This study refines traditional\nDELs by integrating alternative techniques like photocross-linking\nscreening to enhance chemical diversity. Combining these methods improved\npredictive performance for small molecule identification models. Using\nthis approach, we predicted active small molecules for BRD4 and p300,\nachieving hit rates of 26.7 and 35.7%. Notably, the identified compounds\nexhibit smaller molecular weights and better modification potential\ncompared to traditional DEL molecules. This research demonstrates\nthe synergy between DEL and AI technologies, enhancing drug discovery.