Figure S4: Ancestry Accuracy Differentials from Deep-Learning Model for Tumor-Type Prediction Using Targeted Clinical Genomic Sequencing Data

(A) Overall proportion of all annotated ancestries across the training set (left) and testing set (right). Numbers over each bar indicate total within each category. EUR, European; ADM, Admixed; EAS, East Asian; AFR, African; SAS, South Asian; NAM, Native American (B) High-confidence proportion (left) and accuracy (right) for European (EUR), East Asian (EAS), African (AFR) and South Asian (SAS) compared to the overall high-confidence proportion and accuracy across the test set. P-values from a two-sided Fisher's exact test comparing the proportions of these metrics to the overall distribution per ancestry are shown above each bar.