Open AccessA pan‐cancer integrative pathway analysis of multi‐omics data
Author(s) -
Linder Henry,
Zhang Yuping
Publication year - 2020
Publication title -
quantitative biology
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.707
H-Index - 15
eISSN - 2095-4697
pISSN - 2095-4689
DOI - 10.1007/s40484-019-0185-6
Subject(s) - pathway analysis , computational biology , kegg , biology , epigenetics , genomics , systems biology , biological pathway , data science , transcriptome , bioinformatics , computer science , genome , genetics , gene , gene expression
Background Multi‐view ‐omics datasets offer rich opportunities for integrative analysis across genomic, transcriptomic, and epigenetic data platforms. Statistical methods are needed to rigorously implement current research on functional biology, matching the complex dynamics of systems genomic datasets. Methods We apply imputation for missing data and a structural, graph‐theoretic pathway model to a dataset of 22 cancers across 173 signaling pathways. Our pathway model integrates multiple data platforms, and we test for differential activation between cancerous tumor and healthy tissue populations. Results Our pathway analysis reveals significant disturbance in signaling pathways that are known to relate to oncogenesis. We identify several pathways that suggest new research directions, including the Trk signaling and focal adhesion kinase activation pathways in sarcoma. Conclusions Our integrative analysis confirms contemporary research findings, which supports the validity of our findings. We implement an interactive data visualization for exploration of the pathway analyses, which is available online for public access.