Training distribution determines the ceiling of drug-blind cancer sensitivity prediction
For precision oncology researchers, the paper identifies a fundamental flaw in evaluation metrics and proposes training strategies to improve drug-blind sensitivity prediction.
The paper shows that stagnation in drug-blind cancer sensitivity prediction is a metric artifact: global Pearson r is dominated by between-drug differences captured by a trivial drug-mean predictor. Per-drug Pearson r reveals no improvement from drug encodings over cell-only features, and mechanism-stratified training recovers predictive gains.
Precision oncology requires predicting which drugs will suppress a specific tumor from its molecular profile, but drug-blind sensitivity prediction has plateaued despite increasingly complex drug representations. Here we show that this stagnation reflects a metric artifact rather than a representational bottleneck. The standard benchmark, global Pearson r, is dominated by between-drug potency differences that a trivial drug-mean predictor captures without any cell-specific learning. Per-drug Pearson r, which isolates within-drug cell ranking, reveals that no drug encoding improves over cell-only features across four independent datasets. A controlled experiment channeling mechanism-of-action identity as either a drug feature or a training-distribution constraint identifies the cause. Supplying MoA as a feature yields negligible benefit, whereas using it to stratify training raises per-drug r substantially for targeted kinase inhibitors, because pan-cancer co-training suppresses pathway-specific sensitivity signals. Mechanism-stratified training and response matching from pilot observations provide two deployable strategies that together recover the principal sources of predictive gain in drug-blind sensitivity prediction.