Biologically inspired survival analysis based on integrating gene expression as mediator with genomic variants

Ibrahim Youssef, Robert Clarke, Ie Ming Shih, Yue Wang, Guoqiang Yu

Research output: Contribution to journalArticle

Abstract

Accurately linking cancer molecular profiling with survival can lead to improvements in the clinical management of cancer. However, existing survival analysis relies on statistical evidence from a single level of data, without paying much attention to the integration of interacting multi-level data and the underlying biology. Advances in genomic techniques provide unprecedented power of characterizing the cancer tissue in a more complete manner than before, offering the opportunity to design biologically informed and integrative approaches for survival data analysis. Human cancer is characterized by somatic copy number alternation and unique gene expression profiles. However, it remains largely unclear how to integrate the gene expression and genetic variant data to achieve a better prediction of patient survival and an improved understanding of disease progression. Consistent with the biological hierarchy from DNA to RNA, we prioritize each survival-relevant feature with two separate scores, predictive and mechanistic. For mRNA expression levels, predictive features are those mRNAs whose variation in expression levels is associated with survival outcome, and mechanistic features are those mRNAs whose variation in expression levels is associated with genomic variants. Further, we simultaneously integrate information from both the predictive model and the mechanistic model through our new approach, GEMPS (Gene Expression as a Mediator for Predicting Survival). Applied on two cancer types (ovarian and glioblastoma multiforme), our method achieved better prediction power (p-value: 6.18E−03–5.15E−11) than peer methods (GE.CNAs and GE.CNAs. Lasso). Gene set enrichment analysis confirms that the genes utilized for the final survival analysis are biologically important and relevant.

Original languageEnglish (US)
Pages (from-to)231-239
Number of pages9
JournalComputers in Biology and Medicine
Volume77
DOIs
StatePublished - Oct 1 2016

Keywords

  • Copy number alterations
  • Gene expression
  • Genomic data integration
  • Survival analysis
  • Survival feature selection

ASJC Scopus subject areas

  • Computer Science Applications
  • Health Informatics

Fingerprint Dive into the research topics of 'Biologically inspired survival analysis based on integrating gene expression as mediator with genomic variants'. Together they form a unique fingerprint.

  • Cite this