Multi-omics Approach of Risk Stratification for Patients With de Novo Acute Myeloid Leukemia

The investigators will use machine learning to identify features on bone marrow smears and select features that are related to gene mutations, gene expression, or prognosis. The investigators will then use genome-wide transcriptomic profiling to investigate gene expression that is associated with patients' outcomes. The investigators will design a next-generation sequencing panel with unique molecular index and assess its feasibility and robustness in detecting measurable residual disease and optimize the panel/platform/bioinformatic pipeline. Finally, The investigators will use machine learning to integrate bone marrow smear features, gene mutations, gene expression, and measurable residual disease to construct a comprehensive risk assessment system that is based on multi-omics data. The investigators believe that such a platform will help physicians to design the most appropriate treatment strategies for individual patients, not only advancing the concept of precision medicine but also improving patients' prognoses.