EMitool: Explainable Multi-Omics Integration for Disease Subtyping DOI Open Access
Yong Xu, Jun Wu, Chen Chen

и другие.

International Journal of Molecular Sciences, Год журнала: 2025, Номер 26(9), С. 4268 - 4268

Опубликована: Апрель 30, 2025

Disease subtyping is essential for personalized medicine, enabling tailored treatment strategies based on disease heterogeneity. Advances in high-throughput technologies have led to the rapid accumulation of multi-omics data, driving development integration methods comprehensive subtyping. However, existing approaches often lack explainability and fail establish clear links between subtypes clinical outcomes. To address these challenges, we developed EMitool, an explainable tool that leverages a network-based fusion strategy achieve biologically clinically relevant without requiring prior information. Using data from 31 cancer types The Cancer Genome Atlas (TCGA), EMitool demonstrated superior accuracy compared eight state-of-the-art methods. It also provides contribution scores different omics types, enhancing interpretability. EMitool-derived exhibited significant associations with overall survival, pathological stage, tumor mutational burden, immune microenvironment characteristics, therapeutic responses. Specifically, kidney renal cell carcinoma (KIRC), identified three distinct varying prognoses, compositions, drug sensitivities. These findings highlight its potential biomarker discovery precision oncology.

Язык: Английский

EMitool: Explainable Multi-Omics Integration for Disease Subtyping DOI Open Access
Yong Xu, Jun Wu, Chen Chen

и другие.

International Journal of Molecular Sciences, Год журнала: 2025, Номер 26(9), С. 4268 - 4268

Опубликована: Апрель 30, 2025

Disease subtyping is essential for personalized medicine, enabling tailored treatment strategies based on disease heterogeneity. Advances in high-throughput technologies have led to the rapid accumulation of multi-omics data, driving development integration methods comprehensive subtyping. However, existing approaches often lack explainability and fail establish clear links between subtypes clinical outcomes. To address these challenges, we developed EMitool, an explainable tool that leverages a network-based fusion strategy achieve biologically clinically relevant without requiring prior information. Using data from 31 cancer types The Cancer Genome Atlas (TCGA), EMitool demonstrated superior accuracy compared eight state-of-the-art methods. It also provides contribution scores different omics types, enhancing interpretability. EMitool-derived exhibited significant associations with overall survival, pathological stage, tumor mutational burden, immune microenvironment characteristics, therapeutic responses. Specifically, kidney renal cell carcinoma (KIRC), identified three distinct varying prognoses, compositions, drug sensitivities. These findings highlight its potential biomarker discovery precision oncology.

Язык: Английский

Процитировано

0