Pseudogene Coexpression Networks Reveal a Robust Prognostic Signature for Pediatric B-ALL Survival

Area of Science:

• Genomics
• Molecular Biology
• Cancer Research

Background:

• Risk classification in B-cell acute lymphoblastic leukemia (B-ALL) is complex, with current genomic methods not fully capturing outcome variability.
• This suggests undiscovered regulatory mechanisms contributing to B-ALL heterogeneity.

Purpose of the Study:

• To explore pseudogene co-expression as a novel layer of molecular heterogeneity in B-ALL.
• To identify new biomarkers for improved patient risk stratification and survival prediction.

Main Methods:

• Analysis of pseudogene co-expression patterns using single-sample co-expression networks in 1,416 B-ALL patients.
• Principal Component Analysis (PCA) for identifying major sources of variability.
• LASSO-based feature selection to derive a predictive signature.

Main Results:

• Pseudogene co-expression patterns explained significant patient variability and enabled stratification into distinct survival clusters.
• A three-interaction signature was derived, effectively predicting patient survival.
• The RPL7P10-RPS3AP36 interaction was identified as a robust biomarker.

Conclusions:

• Pseudogene co-expression represents a previously unrecognized source of molecular heterogeneity in B-ALL.
• This layer harbors promising molecular markers for future diagnostic and prognostic applications in B-ALL.
• The findings enhance our understanding of B-ALL biology and pave the way for improved precision medicine approaches.