Bayesian multi-source regression and monocyte-associated gene expression predict BCL-2 inhibitor resistance in acute myeloid leukemia

摘要

AUCs calculated across patient-derived ex vivo samples (columns) and 87 drugs (rows) common to OHSU (a; n = 338) and FIMM (b; n = 37) datasets. Red values correspond to higher AUC or more sensitive samples, blue are less sensitive, black are filtered, and gray are missing. Standardized AUCs (i.e., with mean zero and standard deviation one across patients) displayed in heatmap. Raw AUCs displayed in top and side panels. General response across drugs (GRD) is mean of raw AUCs for an individual patient over drugs; mean response across patients (MRP) is mean of raw AUCs for an individual drug over patients. Samples ordered by GRD in each dataset. Drugs ordered by MRP in OHSU dataset. One sample displayed per patient, with sample assayed across the highest number of drugs displayed in cases with multiple samples per patient. Class drug class, Diff/epi differentiation/epigenetic, HSP heat shock protein.