Search

Additional file 2: Fig S1. Gene optimization. Fig S2. Histogram of venetoclax IC50 values from the BeatAML dataset. Fig S3. Optimal rank selection for NMF. Fig S4. Association between the BCL2 family and venetoclax response. Fig S5. Contribution of the BCL2 family to signatures. Fig S6. Identification of BCL2 family-based subtypes in other hematologic malignancies. Fig S7. Mutation status of BCL2 family-based subtypes. Fig S8. Comparison of BCL2 family-based acute myeloid leukaemia (AML) subtypes based on drug response. Fig S9. Assignment of BFSig subtypes in Tavor Dataset. Fig S10. Batch effect correction. Fig S11. External validation of BCL2 family signature-based classifier. Fig S12. Profile of BCL2 family signatures in acute myeloid leukaemia (AML) cell lines. Fig S13. Prediction response to BCL2 family inhibitors in cell lines. Fig S14. Improvement of Prediction Power via Gene Optimization and Rank Selection. Fig S15. Prediction Power using Optimized Genes in External Datasets. Fig S16. Venetoclax Response in NanoString Samples. Fig S17. Relationship between BFL1/MCL1 signature and FAB classification of AML. Fig S18. Correlation between BCL2 family signatures and monocyte signatures. Fig S19. Scheme of gene optimization algorithm. Fig S20. Performance Measurements in Imputation of BCL2 family Profiles.

Additional file 1: Figs. S1–S13.

Additional file 1: Supplementary methods.