The abstract is read at the publisher; the summary is JClub's.
dETM mapped catalytic rates across an adaptive landscape of an imine reductase and found that machine learning could not predict the winning epistatic trajectories.
- Why it matters: Directed evolution remains hard to control because adaptive landscapes and long-term epistasis effects are difficult to analyze with standard sequence-function mapping.
- What they did: The method tracks winners across multi-round evolutionary campaigns, using Deep Evolutionary Trajectory Mapping to map catalytic rates and identify founder variants and positive epistasis.
- The result: Permissive selection and ultrahigh-throughput screening help keep multiple evolutionary paths accessible, while the observed winning trajectories remain computationally unpredictable.
The findingWhy it mattersWhat they didThe result
AI summary of the abstract · may contain errors · not medical advice · read the paper.