Recently, a research team led by Professor Tan Ninghua and Associate Researcher Wang Yinyin from the School of Traditional Chinese Pharmacy at our university collaborated with a team led by Associate Professor Tang Jing from the University of Helsinki in Finland. Their latest research findings, titled “A Network-Driven Framework for Drug Response Precision Prediction of Acute Myeloid Leukemia,” were published in the top-tier journal Advanced Science. Our university is the first corresponding institution for the paper. Associate Researcher Wang Yinyin and Liu Rui, a master's student from the Class of 2023 at our university, are the co-first authors. Associate Professor Tang Jing and Professor Tan Ninghua are the co-corresponding authors.
Acute myeloid leukemia (AML) is a malignant clonal disease originating from hematopoietic stem cells in the bone marrow. Patients exhibit significant variability in response to existing treatment regimens, necessitating the development of personalized drug regimens tailored to individual patient differences. However, the heterogeneity of tumor cells makes it exceptionally challenging to identify reliable predictive biomarkers. Traditional models based on bulk RNA-Seq and in vitro experiments often fail to capture the complex molecular pathways and gene networks underlying treatment response and drug resistance. To address this, the research team developed a network-driven precision drug sensitivity prediction platform named NetAML. The key findings of this study are as follows:
1) Large-scale predictive model construction: Using RNA sequencing data and in vitro drug response data from 520 acute myeloid leukemia patients, the NetAML system developed personalized predictive models for 87 clinical drugs.
2) Integration of network analysis and machine learning: The platform's core lies in combining network analysis and machine learning technologies. This approach transcends the limitations of individual genes to analyze the interaction networks between genes.
3) Revealing the biological mechanisms driving drug responses: NetAML successfully identified gene signatures with high biological significance. These signatures are not isolated gene sets but reflect the complex molecular interaction networks driving different drug responses.
4) Discovery of new resistance mechanisms: By analyzing the gene signature patterns of the model, the co-expression of the C19ORF59 gene and the FLT3 gene was revealed to be significantly associated with patient resistance to FLT3 inhibitor drugs. This discovery provides new insights into understanding resistance mechanisms and developing strategies to overcome them.
5) Advancing personalized therapy: By constructing drug-specific predictive models, identifying clinically applicable biomarkers, and predicting specific patients' sensitivity to different drugs, this research provides a reference for developing highly personalized treatment plans.
Diagram
The NetAML platform converts complex genomic, in vitro pharmacodynamic, and other multidimensional data into predictive tools with clear clinical guidance, demonstrating the enormous potential of AI-driven big data analysis. This work was selected as the cover article for the current issue.
The above work was supported by the National Natural Science Foundation of China Youth Project (No. 82405199), Jiangsu Province Youth Project (BK20231024), and China Pharmaceutical University Talent Introduction Project.
Article link: https://doi.org/10.1002/advs.202506447
