Combining lipidomics and machine learning to identify lipid biomarkers for nonsyndromic cleft lip with palate

• DOI: 10.1172/jci.insight.186629
• Publication Date: 2025-05-07T18:00:36Z
• Authors: Shanshan Jia; Weidong Xie; Chunqing Yang; Yizhang Dong; Wenting Luo; Hui Gu; Xiaowei Wei; Wei Ma; Dan Liu; Songying Cao; Yuzuo Bai; Wei Li; Zhengwei Yuan

Nonsyndromic cleft lip with palate (nsCLP) is a common birth defect disease. Current diagnostic methods comprise fetal ultrasound images, which are mainly limited by position and technician skills. We aimed to identify reliable maternal serum lipid biomarkers diagnose nsCLP. Eight-feature selection were used assess the dysregulated lipids from untargeted lipidomics in discovery cohort. The robust rank aggregation algorithm was applied on these selected lipids. data subsequently processed using 7 classification models retrieve panel of 35 candidate biomarkers. Potential evaluated targeted validation Seven multivariate analyses constructed for model achieved high performance 3 determining A showed great potential nsCLP diagnosis. FA (20:4) LPC (18:0) also significantly downregulated early samples group additional demonstrate applicability robustness machine-learning analyze lipidomic efficient biomarker screening.