Related Experiment Video
Updated: Sep 10, 2025
13:19
Microarray-based Identification of Individual HERV Loci Expression: Application to Biomarker Discovery in Prostate Cancer
Published on: November 2, 2013
16.7K
Machine Learning-Based Gene Expression Analysis to Identify Prognostic Biomarkers in Upper Tract Urothelial
Bernat Padullés1,2, Ruben López-Aladid3, Mercedes Ingelmo-Torres1,2
1Urology Department and Laboratory, Hospital Clínic de Barcelona, 08036 Barcelona, Spain.
Cancers
|August 28, 2025
Summary
This study identifies ten key genes using machine learning to predict upper tract urothelial carcinoma (UTUC) progression. These gene expression biomarkers show high discriminative ability, potentially improving risk stratification for this rare cancer.
Area of Science:
- Oncology
- Genomics
- Bioinformatics
Background:
- Upper tract urothelial carcinoma (UTUC) is a rare, aggressive cancer with limited prognostic tools.
- The molecular drivers of UTUC are poorly understood due to its low incidence.
- Improved risk stratification is needed for effective postoperative management.
Purpose of the Study:
- To identify gene expression biomarkers for predicting UTUC progression.
- To explore the potential of machine learning in analyzing UTUC transcriptomic data.
- To enhance risk stratification and guide treatment decisions for UTUC patients.
Main Methods:
- Applied machine learning to gene expression data from 17 pT2/pT3 UTUC radical nephroureterectomy (RNU) specimens.
- Utilized RNA sequencing and differential gene expression analysis (DESeq2).
- Employed logistic regression and ROC analysis for predictive power evaluation.
Main Results:
- Identified 76 differentially expressed genes between progressive and non-progressive UTUC.
- A random forest classifier pinpointed ten key genes with prognostic potential.
- Achieved an AUC of 0.88, demonstrating high discriminative ability for these genes.
Conclusions:
- Machine learning integrated with transcriptomic analysis shows promise for identifying UTUC prognostic biomarkers.
- The identified genes are linked to immune regulation, cell cycle, and tumor progression.
- Further validation in larger cohorts is essential for clinical application.
