Skip to main content

Table 4 Predictive power on the test set of multi-task and single-task models trained on the COX-1 and COX-2 data sets

From: KekuleScope: prediction of cancer cell line sensitivity and compound potency using convolutional neural networks trained on compound images

Model RMSE ± SD on COX-1 RMSE ± SD on COX-2
Multi-target COX-1 and COX-2 0.73 ± 0.05 0.75 ± 0.05
Single-target COX-1 0.73 ± 0.06 NA
Single-target COX-2 NA 0.72 ± 0.05