Additional file 1 of A machine learning framework develops a DNA replication stress model for predicting clinical outcomes and therapeutic vulnerability in primary prostate cancer

Additional file 1: Table S1. Gene sets of DNA replication stress. Table S2. Hyperparameters used in the machine learning algorithms. Table S3. Collection of prostate cancer signatures. Table S4. Gene lists of pathways. Table S5. The list of curated targetable genes. Table S6. The result of univariate Cox regression analysis of DNA replication stress-related genes. Table S7. The result of Bootstrapping-based Cox regression analysis. Table S8. The result of machine learning benchmark. Table S9. The result of signatures comparison. Table S10. Somatic copy number alterations in RSS-high and RSS-low groups. Table S11. Somatic mutation characteristics in the RSS-high and RSS-low groups. Table S12. Clinicopathologic characteristics of prostate cancers in the included cohorts. Table S13. The result of single sample gene set enrichment analysis in the Meta-cohort. Table S14. The CIBERSORT result of 905 prostate cancer samples. Table S15. The result of Spearman’s rank-order correlation between replication stress signature and druggable genes in the TCGA-PRAD cohort. Table S16. The result of Spearman’s rank-order correlation between replication stress signature and druggable genes in the DKFZ-PRAD cohort. Table S17. The result of a meta-analysis of differential gene analysis. Table S18. The result of CMap analysis. Table S19. Comparison of clinical characteristics among included cohorts.