Researchers have developed DegradeQuery, a novel framework designed to predict the degradation of proteolysis-targeting chimeras (PROTACs). This context-aware system leverages existing, but often label-missing, molecule-target-E3 records by employing a counterfactual tuple pretraining objective. By contrasting actual molecular interactions with hypothetical alternatives, DegradeQuery learns contextual associations without requiring explicit activity labels. The framework achieved a 0.9065 AUC and 0.8500 accuracy on the PROTAC-8K benchmark, surpassing existing methods and demonstrating the utility of incompletely labeled datasets. AI
IMPACT This framework could significantly improve the efficiency of drug discovery by enabling better prediction of PROTAC degradation from limited experimental data.
RANK_REASON The cluster describes a new research paper detailing a novel framework for a specific scientific prediction task. [lever_c_demoted from research: ic=1 ai=1.0]
Read on Hugging Face Daily Papers →
AI-generated summary · Google Gemini · from 2 sources. How we write summaries →
