Answers are grounded in this paper and cite their sources.
Ulcerative colitis (UC) is a chronic inflammatory bowel disease. Lysosomes are known to play critical roles in inflammatory bowel disease (IBD) pathogenesis, but the involvement of lysosome‑related molecules in pediatric UC remains poorly understood. This study was designed to identify key lysosome‑related hub genes and, based on these genes, to develop machine learning models for diagnostic evaluation and prognostic prediction in pediatric UC. Transcriptomic data were obtained from the Gene Expression Omnibus (GEO) database. Candidate hub genes were filtered by differential expression analysis and weighted gene co-expression network analysis (WGCNA), followed by quantitative real-time polymerase chain reaction (qRT-PCR) validation. Four machine learning models, namely random forest (RF), support vector machine (SVM), linear discriminant analysis (LDA), and k-nearest neighbor (KNN), were constructed for diagnostic discrimination of pediatric UC; RF and LDA were further applied to predict colectomy risk within one year of diagnosis. Correlations between the hub genes and disease severity, immune infiltration, intestinal barrier function, and intestinal fibrosis were also analyzed. Three hub genes, namely caudal type homeobox 2 ( CDX2 ), sialic acid acetylesterase ( SIAE ), and myotubularin 1 ( MTM1 ), were markedly downregulated in pediatric UC lesions. Their expression levels were negatively correlated with the Pediatric Ulcerative Colitis Activity Index (PUCAI) score and the total Mayo score, and were lower in children who required colectomy within one year of diagnosis. In the diagnostic machine learning models, area under the curve (AUC) values ranged from 0.990 to 0.998 in the training cohort and from 0.870 to 0.981 in the test cohort, with RF and LDA showing the best stability (test-cohort AUCs = 0.981 and 0.970, respectively). For the risk of colectomy within one year of diagnosis, the RF model achieved AUCs of 0.932 and 0.905 in the training and test cohorts, respectively, while the LDA model achieved 0.850 and 0.930, respectively. Furthermore, these three genes were positively associated with genes related to intestinal epithelial barrier function and negatively correlated with pro-inflammatory and fibrosis-related genes. The lysosome-related genes CDX2 , SIAE , and MTM1 are closely associated with pediatric UC. The three-gene signature may assist in UC diagnosis, help estimate disease severity, and indicate the risk of progression to colectomy, providing novel clues for understanding the molecular landscape and clinical management of pediatric UC. Not applicable.
No takes yet. Share an insight, caveat, or question.
Wu et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: