A Network‐Driven Framework for Drug Response Precision Prediction of Acute Myeloid Leukemia

Acute myeloid leukemia (AML) is a clonal malignancy of myeloid progenitor cells that demonstrates highly variable responses to current regimens, highlighting the need for precision medicine. However, reliable biomarkers for precision medicine treatment remain elusive due to cellular heterogeneity. Conventional Models based on bulk RNA sequencing and ex vivo assays often fail to capture the intricate molecular pathways and gene networks that underlie treatment response and resistance. Here, NetAML, a novel network-based precision medicine platform that systematically develops 87 drug sensitivity prediction models for 87 clinical drugs using ex vivo drug responses from 520 AML patients with RNA-Seq is presented. The approach leverages network-based analysis and machine learning to identify biologically interpretable gene signatures that capture the complex molecular interactions driving differential drug responses. Notably, the signature genes derived from the models reveal distinct cellular mechanisms. For instance, the co-expression of C19ORF59 and FLT3 is associated with resistance to FLT3 inhibitors. In summary, NetAML offers a powerful strategy for personalized AML treatment by constructing drug-specific models, identifying clinically actionable biomarkers, and supporting the development of optimized, patient-specific therapeutic regimens.