Splicing junction-based classifier for the detection of abnormal constitutive activation of the KEAP1-NRF2 system.

Abstract

The KEAP1-NRF2 system plays a crucial role in responding to oxidative and electrophilic stress. Its dysregulation can cause the overexpression of downstream genes, a known cancer hallmark. Understanding and detecting abnormal KEAP1-NRF2 activity is essential for understanding disease mechanisms and identifying therapeutic targets. This study presents an approach that analyzes splicing patterns by a naive Bayes-based classifier to identify constitutive activation of the KEAP1-NRF2 system, focusing on the higher presence of abnormal splicing junctions as a subproduct of overexpression of downstream genes. Our splicing-based classifier demonstrated robust performance, reliably identifying activation of the KEAP1-NRF2 pathway across extensive datasets, including The Cancer Genome Atlas and the Sequence Read Archive. This shows the classifier's potential to analyze hundreds of thousands of transcriptomes, highlighting its utility in broad-scale genomic studies and provides a new perspective on utilizing splicing aberrations caused by overexpression as diagnostic markers, offering potential improvements in diagnosis and treatment strategies.