Multi-method Machine-Learning Identification of IL11 as a Key Gene in Kidney Stone Disease: Mechanistic Clues and...

Randall's plaque (RP) is a key initiating lesion of idiopathic calcium stones at the renal papilla, and transcriptomic profiling of human RP has revealed inflammation- and matrix-remodeling-related changes, including upregulation of interleukin 11 (IL11). However, under a pure kidney-stone phenotype, a rigorous framework combining multi-algorithm consensus screening with fuller diagnostic appraisal (calibration, internal validation, and decision-curve analysis (DCA)) remains needed.

GSE73680 (RP, n=29; normal papilla, n=33) was obtained from the Gene Expression Omnibus (GEO). Exploratory differential expression (limma) and weighted gene co-expression network analysis (WGCNA) were intersected, followed by Gene Ontology (GO)/Kyoto Encyclopedia of Genes and Genomes (KEGG) enrichment and Search Tool for the Retrieval of Interacting Genes/Proteins (STRING) protein-protein interaction (PPI) analysis. Consensus genes were defined by five feature-selection algorithms: least absolute shrinkage and selection operator (LASSO), support vector machine recursive feature elimination (SVM-RFE), random forest, XGBoost, and Boruta. For IL11, we assessed expression, built a penalized logistic nomogram with receiver operating characteristic (ROC), calibration, and patient-blocked internal validation, and performed Cell-type Identification by Estimating Relative Subsets of RNA Transcripts (CIBERSORT) deconvolution and correlation-ranked KEGG gene set enrichment analysis (GSEA).

The intersection of 39 exploratory differentially expressed genes (DEGs) with the WGCNA green module yielded 23 candidates, of which IL11 was the sole five-method consensus gene and was significantly upregulated in RP (Wilcoxon P=8.9×10⁻⁵). The single-gene model achieved an apparent area under the curve (AUC) of 0.782 (95% confidence interval (CI) 0.663-0.901) and an out-of-bag (OOB) AUC of 0.756. PPI linked IL11 to SPP1 and TNFRSF11B, and GSEA enriched lysosome, cell adhesion molecules, and leukocyte transendothelial migration pathways.

Building on prior studies, this work strengthens the evidence chain for IL11 as a priority target under a pure kidney-stone/RP phenotype through five-algorithm consensus, complete diagnostic appraisal, and adhesion/migration-related mechanistic clues.