Researchers have developed GenFish, a novel generative retrieval approach for reverse target fishing, a crucial process in drug discovery. This method utilizes residual vector quantization to convert protein language model features into shared semantic identifiers, enabling the autoregressive modeling of codes conditioned on query information. GenFish outperforms existing discriminative neural and non-neural baselines on ProTargetBench, a new proteome-scale benchmark designed to evaluate full-catalogue recovery of experimentally recorded positives. The semantic identifiers generated by GenFish also demonstrate the ability to capture biologically meaningful relationships among proteins. AI
IMPACT This generative retrieval approach could accelerate drug discovery by improving the identification of protein targets for molecules.
RANK_REASON The cluster contains an academic paper detailing a new method and benchmark for a specific scientific task. [lever_c_demoted from research: ic=1 ai=1.0]
AI-generated summary · Google Gemini · from 1 sources. How we write summaries →
