Systematic Characterization of Splicing Dysregulation in Pan Solid Tumor Transcriptome

Area of Science:

• Genomics
• Molecular Biology
• Computational Biology

Background:

• Splicing dysregulation, often from spliceosomal mutations, drives disease and treatment resistance, particularly in hematologic cancers.
• While less common in solid tumors, splicing disorders are pervasive and contribute to tumorigenesis.
• A systematic understanding of splicing dysregulation patterns across solid tumors is lacking.

Purpose of the Study:

• To computationally uncover the pan-cancer splicing dysregulation landscape.
• To identify joint modular patterns of splicing factors (SFs) and alternative splicing events (ASEs).
• To characterize the full spectrum of splicing dysregulation patterns across 31 human solid tumors.

Main Methods:

• Development of a computational method, SMNPLS (Sparse Multi-Network Regularized Partial Least Squares).
• Extraction of joint modular patterns from paired SF expression and ASE matrices.
• Analysis of splicing dysregulation patterns across 40% of TCGA solid tumors.

Main Results:

• Six unique ASE-SF co-module patterns were identified, involving 1,570 ASEs and 170 SFs.
• Common splicing dysregulation patterns were observed in digestive, renal, and urogenital tumors.
• Brain tumors exhibited a distinct splicing pattern with high ASE-SF correlation, and novel potentially oncogenic regulatory relationships were identified.

Conclusions:

• The study characterizes the complete spectrum of splicing dysregulation in solid tumors.
• It highlights both similarities and specificities in splicing-derived pathogenesis across different cancer types.
• Findings provide insights into splicing-driven tumorigenesis and potential therapeutic targets.