For global researchers and biotech investors, this study signals a new frontier: machine learning can untangle complex, co-morbid conditions and identify biomarkers that may guide both diagnosis and drug repurposing—an approach that scales far beyond the obesity clinic.
Chinese scientists have employed machine learning to identify three co-morbid biomarkers for sarcopenic obesity (SO)—a debilitating condition that traps older adults in a cycle of muscle loss and fat accumulation, sharply reducing quality of life. Using data from the China Health and Retirement Longitudinal Study (CHARLS), the team found that SO incidence in China surged from 16.1% in 2011 to 20.4% in 2018, underscoring a growing public health challenge.
The researchers applied a cascade of machine learning algorithms—LASSO, XGBoost, SVM-REF, and Random Forest—to winnow a vast pool of differentially expressed genes and gut-microbiota metabolite targets down to three key genes: ALDH1A3, CSF1R, and PHGDH. These candidates were then validated across multiple independent datasets, demonstrating robust diagnostic performance with AUC values exceeding 0.72. Further validation came from single-cell sequencing and immunohistochemistry in a high-fat-diet mouse model, where ALDH1A3 and CSF1R were significantly upregulated in muscle tissue.
Beyond diagnosis, the study explored therapeutic potential. Immune infiltration analysis revealed a notable increase in resting NK cells in both obesity and sarcopenia states. Crucially, molecular docking simulations identified Birinapant—a compound already under investigation in oncology—as stably binding to the key gene targets. This suggests a possible path to drug repurposing for a condition that currently lacks specific pharmacological therapy.
Why it matters:
This research demonstrates how machine learning can integrate multi-omics data to identify clinically actionable biomarkers for complex, co-morbid conditions. For drug developers and clinicians, the identification of Birinapant as a potential therapeutic candidate opens a fast-track repurposing opportunity. For China’s aging population, where SO is rising sharply, these findings offer a pathway toward earlier diagnosis and targeted intervention.
ScientificChina — tracking what’s happening in Chinese science, technology, research, and industrial innovation in a way global professionals can actually use.
Follow ScientificChina for deeper insight into China’s evolving science, technology, and industrial landscape.
No products in the cart.