Population effect model identifies gene expression predictors of survival outcomes in lung adenocarcinoma for both Caucasian and Asian patients.

ABSTRACT:

Background

We analyzed and integrated transcriptome data from two large studies of lung adenocarcinomas on distinct populations. Our goal was to investigate the variable gene expression alterations between paired tumor-normal tissues and prospectively identify those alterations that can reliably predict lung disease related outcomes across populations.

Methods

We developed a mixed model that combined the paired tumor-normal RNA-seq from two populations. Alterations in gene expression common to both populations were detected and validated in two independent DNA microarray datasets. A 10-gene prognosis signature was developed through a l1 penalized regression approach and its prognostic value was evaluated in a third independent microarray cohort.

Results

Deregulation o

SUBMITTER: Cai G 

PROVIDER: S-EPMC5398559 | biostudies-literature | 2017

REPOSITORIES: biostudies-literature