Fig. 2.

Performance comparison between multitask learning and single-task learning approaches using JUN family TFs. Models are trained with datasets from (a) JUND across multiple cell types, and (b) multiple TFs in the JUN family across multiple cell types. MTL, multitask learning; MTL-sampled, multitask learning training data that has been subsampled to match the number of samples in the corresponding single-task models; STL, single-task learning. The right panels in (a) and (b) are the averaged auPRC of the models shown in the corresponding left panels. Error bars represent standard error of the mean across all training conditions. P-values are calculated using the Wilcoxon signed-rank test using auPRC scores across all conditions