Sign in Subscribe
Cancer research

Multimodal Gene Expression and Methylation Profiling Reveals Misclassified Tumors Beyond Histological Diagnosis

Strategic Insights

— 1 min read

Investigators at Thompson Rivers University have published a new report on Cancer, highlighting the importance of integrating gene expression and DNA methylation data to improve tumor classification accuracy and detect misclassified tumors.

The research, supported by the Natural Sciences and Engineering Research Council of Canada (NSERC), utilized 6216 samples from The Cancer Genome Atlas (TCGA) to evaluate classification accuracy across single and integrated platforms. The results showed that mRNA and methylation data alone achieved accuracies of 97% and 95.4%, respectively. However, the integration of these data further reduced false positives and improved the identification of outliers, including histologically misclassified cases such as papillary renal cell carcinoma samples clustering with bladder cancer.

Key Takeaways:

  • The study highlights the value of combining genetic and epigenetic profiles to refine cancer diagnostics, with an integrated approach achieving an accuracy rate of 97.4% compared to 95.4% for mRNA alone and 97% for methylation alone.
  • The research identified molecular subtypes correlated with somatic mutations and patient survival, offering clinically relevant insights.
  • The study demonstrates the potential of multimodal gene expression and methylation profiling to diagnose and manage cancer more effectively.
  • The use of Support Vector Machines (SVMs) and hierarchical clustering improved classification accuracy and reduced false positives.
  • The research provides a scalable quality control tool for molecular oncology.
  • The study suggests that histological diagnosis alone may overlook key molecular differences or result in misclassification.

Statistics:

  • 97.4% accuracy rate achieved by integrating gene expression and DNA methylation data.
  • 95.4% accuracy rate achieved by methylation data alone.
  • 97% accuracy rate achieved by mRNA data alone.
  • 6216 samples used from The Cancer Genome Atlas (TCGA) for the study.

Sources:

  • NewsRx. Studies from Thompson Rivers University Yield New Information about Cancer (Multimodal Gene Expression and Methylation Profiling Reveals Misclassified Tumors Beyond Histological Diagnosis). Health & Medicine Week. July 25, 2025; p 4619.
  • Multimodal Gene Expression and Methylation Profiling Reveals Misclassified Tumors Beyond Histological Diagnosis. Electronics, 2025;14(12).

Read more