@mikejg84 on Backlist

Deep-learning gene-perturbation predictors still trail linear baselines

A Nature Methods paper finds deep-learning approaches to gene perturbation effect prediction do not yet beat simple linear baselines