Transcriptomic analysis reveals novel targets in benign schwannoma using machine learning.

Abstract

Background & objective: This study aimed to identify key immune-related biomarkers of benign schwannoma through machine learning-assisted transcriptomic and single-cell analyses, and to construct a predictive model for disease evaluation.

Methods: Transcriptomic data from the GSE108524 dataset were utilized for immune subtyping and immune cell infiltration analysis. Key biomarkers were screened using the Least Absolute Shrinkage and Selection Operator (LASSO), Support Vector Machine (SVM), and Random Forest algorithms. A nomogram-based predictive model was developed (area under the curve [AUC] = 0.67) and evaluated using accuracy, sensitivity, specificity, and F1-score metrics. The distribution of identified biomarkers across immune cell subsets was validated using scRNA-seq, with a particular focus on T cells and macrophages. Functional roles of ANGPTL1, IL17RC, LTBR, OLR1, and TGFBR1 were further verified through in vitro assays and in vivo using an NF2-knockout mouse model.

Results: Five immune-related biomarkers were identified. Among them, ANGPTL1 and IL17RC inhibited tumor cell proliferation and migration, whereas LTBR, OLR1, and TGFBR1 promoted these processes. These genes exhibited differential expression across immune subtypes and were enriched in tumor-associated immune cells. Both in vitro and in vivo experiments substantiated their biological significance in schwannoma progression.

Conclusion: This study identifies five novel immune-related biomarkers with functional relevance in benign schwannoma, providing new insights into its immune microenvironment and tumor biology. The predictive model offers a foundation for risk stratification and personalized therapeutic strategies. These findings complement known markers such as NF2, SOX10, and S100B, highlighting their potential translational value as diagnostic and therapeutic targets.