Using Machine Learning Framework To Identify Prognostic Biomarkers In Neuroblastoma

Neuroblastoma is nan astir communal coagulated tumor successful infants and accounts for astir 15% of each pediatric cancer-related deaths. Despite decades of advancement successful surgery, chemotherapy, and stem compartment therapies, endurance for high-risk patients remains nether 60%. Current biomarkers—such arsenic MYCN amplification aliases ALK mutations—offer constricted reach, coming only successful subsets of patients aliases requiring analyzable testing. These limitations time off a captious spread successful efficaciously predicting illness progression and guiding treatment. Due to these challenges, location is simply a pressing request to uncover new, interpretable biomarkers that tin amended early consequence stratification and thrust guardant much personalized therapies.

A investigation squad astatine nan Children's Hospital of Chongqing Medical University has unveiled a powerful instrumentality learning model for identifying prognostic biomarkers successful neuroblastoma. Published (DOI: 10.1002/pdi3.70009) successful Pediatric Discovery successful May 2025, nan study leverages bulk and single-cell RNA sequencing information from complete 1,200 patients to build a broad prognostic network. The team's integrative attack not only isolated 11 cardinal cistron signatures but besides revealed really these genes interact pinch nan tumor microenvironment and supplier responses—paving nan measurement for much precise and effective curen plans successful pediatric oncology.

To decipher neuroblastoma's analyzable familial architecture, nan researchers applied an enhanced type of nan stSVM instrumentality learning exemplary to analyse bulk RNA-seq information from 1,207 patients. This process uncovered 528 genes powerfully linked to endurance outcomes. Using weighted cistron co-expression web study (WGCNA), nan squad filtered this database to 11 hub genes—AURKA, BLM, BRCA1, BRCA2, CCNA2, CHEK1, E2F1, MAD2L1, PLK1, RAD51, and notably, RFC3. High look of RFC3 correlated pinch mediocre prognosis and debased earthy slayer (NK) compartment activity, hinting astatine its domiciled successful immune evasion. The study besides revealed that tumors pinch elevated RFC3 look were much delicate to vincristine and cyclophosphamide—standard chemotherapy agents. Further exploration utilizing single-cell RNA sequencing confirmed higher RFC3 look successful epithelial and myeloid cells among short-survival patients, on pinch reduced T compartment infiltration. These multilayered findings not only item RFC3 arsenic a caller biomarker but besides propose it whitethorn style nan immune scenery and supplier consequence successful neuroblastoma. By combining cistron networks, immune signatures, and supplier sensitivity profiles, nan investigation offers a rich, systems-level knowing of nan disease.

Source:

Journal reference:

Tang, S., et al. (2025). Identification of Prognostic Biomarkers successful Gene Expression Profile of Neuroblastoma Via Machine Learning. Pediatric Discovery. doi.org/10.1002/pdi3.70009.