Abstract 7346: Optimizing pan-cancer driver pathway analysis from patient transcriptomics and phosphoproteomics

Abstract Alterations in cell signaling pathways drive cancer progression. Such changes can be detected by transcriptomics and proteomics profiling of tumor samples. Although proteomics profiling captures these alterations more directly, RNA-seq transcriptomics remains the most widely available and cost-effective source of data for elucidating tumor-specific cell signaling mechanisms. The eXpression2Kinases (X2K) pipeline is a computational workflow that starts with differentially expressed mRNAs, the identified differentially expressed genes are then used as input for transcription factor enrichment analysis. Then, the top-ranked transcription factors are connected via additional protein-protein interactions from several databases. Finally, the top-ranked transcription factors and the proteins that directly interact and connect these factors are subjected to kinase enrichment analysis. Such analysis identifies protein kinases most likely responsible for the observed changes in gene expression. The X2K pipeline can be applied to analyze tumor-specific transcriptomics to infer upstream regulatory transcription factors, protein intermediates, and kinases that are likely the drivers of cancer progression. While the X2K pipeline infers regulatory pathways from transcriptomics alone, phosphoproteomics can be used to independently validate the inferred kinases, and to calibrate the parameters of the X2K pipeline. Furthermore, this multiomics approach predicts cell signaling pathways with multiple layers of evidence by linking changes observed in the phosphoproteome to changes observed in the transcriptome. Here, we used the transcriptomics and phosphoproteomics profiles from the Clinical Proteomic Tumor Analysis Consortium (CPTAC) pan-cancer dataset, including 1,020 samples across 10 different cancer types, to optimize the parameters of the X2K pipeline and predict pan-cancer as well as tumor-specific signaling pathways. While inferred pathways were distinct between different tumor types, we observed global enrichment of receptor tyrosine kinases and cell cycle kinases. These pathways were subsequently tested for association with patient survival to identify pathways that may be effective therapeutic targets, for example, PTK2 signaling was identified for many lung cancer tumors (LUAD and LSCC) as well as some breast, head & neck squamous. Altogether, the X2K pipeline presents a rational approach to better identify driver regulatory mechanisms from patient-specific profiling of tumors with both transcriptomics and phosphoproteomics. Citation Format: Eden Z. Deng, Giacomo B. Marino, Daniel J. Clarke, Weiping Ma, Pei Wang, Avi Ma'ayan. Optimizing pan-cancer driver pathway analysis from patient transcriptomics and phosphoproteomics [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2024; Part 1 (Regular Abstracts); 2024 Apr 5-10; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2024;84(6_Suppl):Abstract nr 7346.