Adipose precursor cell-associated markgene combined with single-cell sequencing with machine learning to build an obesity prediction model

Abstract Objective: The purpose of this study is to Identify genes associated with adipose precursor cells(APCs), and apply machine learning methods, such as the minimum absolute contraction and selection operator (LASSO) and SVM recursive feature elimination (SVM-rfe), to find large obesity markers. Method: We obtained a single-cell RNA sequencing dataset GSE133486 and three APCs RNA chip datasets GSE 24883, GSE25401 and GSE156906 from GEO.The R package was applied to find common differential genes, and key genes were obtained by LASSO with SVM-ref. Results: THBS2, VASN, VCAN, ALDH1A3,GLUL, SLC4A4 take participate in the formation of APCs. Conclusion: This study found that changes in different genes can regulate the occurrence and development of obesity, and discusses the possibility of these genes in the clinic.