Unleashing the potential of exosome ncRNAs for early gastric cancer detection—a critical appraisal of machine learning approaches

We read with great interest the article by Cai et al ,1 titled ‘Construction of exosome non-coding RNA feature for non-invasive, early detection of gastric cancer patients by machine learning: a multi-cohort study’, published in Gut . The study presents a novel approach for the early detection of gastric cancer (GC) using serum exosome non-coding RNAs (ncRNAs) and machine learning, which is a significant step forward in the field of liquid biopsy for cancer diagnostics. The study by Cai et al has several notable strengths. First, the comprehensive multi-cohort design, including both training and external validation cohorts, provides robust evidence for the diagnostic potential of the identified exosome ncRNA feature. The use of machine learning algorithms, particularly LASSO-logistic regression, to develop the combined diagnostic model (cd-score) is innovative and demonstrates high diagnostic accuracy with an area under the curve of 0.959 in the training cohort and 0.949 in the external validation cohort. Additionally, the study highlights the potential of DGCR9 as a therapeutic target, supported by in vitro and in vivo experiments. Despite these strengths, …