New benchmark and model tackle enzymatic reaction prediction

Researchers have introduced VenusRX-Bench, a new benchmark designed to evaluate models on enzymatic reaction prediction, retrosynthesis, and Enzyme Commission (EC) number prediction. This benchmark highlights a significant gap between chemical and enzymatic models, attributed to limited data and the complexity of modeling biomolecular structures. To address this, the team developed VenusRX, a T5-based sequence-to-sequence model that jointly learns these tasks, incorporating EC conditioning and constrained decoding for improved biomolecule generation. VenusRX demonstrates competitive performance across various tasks, underscoring the bidirectional relationship between reaction structure and catalytic function. AI

IMPACT This research could advance AI's capabilities in understanding and predicting complex biochemical processes, potentially aiding drug discovery and synthetic biology.

RANK_REASON The cluster describes a new academic paper introducing a benchmark and a model for enzymatic reaction prediction. [lever_c_demoted from research: ic=1 ai=1.0]

Read on arXiv cs.AI →

• alphaXiv
• arXiv
• CatalyzeX
• DagsHub
• Enzyme Commission number
• Gotit.pub
• Hugging Face
• ScienceCast
• T5 Text To Text Transfer Transformer
• VenusRX
• VenusRX-Bench

• paper
• model release
• other

AI-generated summary · Google Gemini · from 1 sources. How we write summaries →