Researchers Train ML System Predicts Enantioselectivity
Researchers at the University of Utah and UCLA publish an accelerated preview in Nature on Feb. 11, 2026, describing an ML workflow that predicts enantioselectivity in nickel-based asymmetric cross-coupling reactions using data from four prior studies. The model accurately forecasts product handedness with limited experimental input, cutting typical screening from roughly 50–60 reactions to about 5–10, reducing time and cost for reaction optimization and pharmaceutical scale-up.
Key Points
- 1Demonstrate ML model predicts enantioselectivity in nickel-based asymmetric cross-coupling using limited experimental data
- 2Reduce experimental workload dramatically, cutting required reactions from ~50–60 to roughly 5–10, saving time and cost
- 3Enable chemists and pharma to prioritize ligand and substrate choices, accelerating route optimization and scale-up decisions
Scoring Rationale
High methodological novelty and strong experimental validation, limited to current nickel-based datasets though transferable in principle
Sources
Public references used for this report.
Practice interview problems based on real data
1,625 SQL & Python problems across 15 industry datasets — the exact type of data you work with.
Try 250 free problems
