Research Papers:

Genome-wide screen identifies a novel prognostic signature for breast cancer survival

Xuan Y. Mao, Matthew J. Lee, Jeffrey Zhu, Carissa Zhu, Sindy M. Law and Antoine M. Snijders _

PDF  |  HTML  |  Supplementary Files  |  How to cite

Oncotarget. 2017; 8:14003-14016. https://doi.org/10.18632/oncotarget.14776

Metrics: PDF 3666 views  |   HTML 2198 views  |   ?  


Xuan Y. Mao1, Matthew J. Lee1, Jeffrey Zhu1, Carissa Zhu1, Sindy M. Law2, Antoine M. Snijders1

1Biological Systems and Engineering Division, Lawrence Berkeley National Laboratory, Berkeley, California, USA

2Department of Psychiatry, Weill Institute for Neurosciences, University of California San Francisco, San Francisco, California, USA

Correspondence to:

Antoine M. Snijders, email: [email protected]

Keywords: breast cancer, prognostic score, relapse-free survival, gene biomarkers

Received: August 15, 2016     Accepted: December 31, 2016     Published: January 21, 2017


Large genomic datasets in combination with clinical data can be used as an unbiased tool to identify genes important in patient survival and discover potential therapeutic targets. We used a genome-wide screen to identify 587 genes significantly and robustly deregulated across four independent breast cancer (BC) datasets compared to normal breast tissue. Gene expression of 381 genes was significantly associated with relapse-free survival (RFS) in BC patients. We used a gene co-expression network approach to visualize the genetic architecture in normal breast and BCs. In normal breast tissue, co-expression cliques were identified enriched for cell cycle, gene transcription, cell adhesion, cytoskeletal organization and metabolism. In contrast, in BC, only two major co-expression cliques were identified enriched for cell cycle-related processes or blood vessel development, cell adhesion and mammary gland development processes. Interestingly, gene expression levels of 7 genes were found to be negatively correlated with many cell cycle related genes, highlighting these genes as potential tumor suppressors and novel therapeutic targets. A forward-conditional Cox regression analysis was used to identify a 12-gene signature associated with RFS. A prognostic scoring system was created based on the 12-gene signature. This scoring system robustly predicted BC patient RFS in 60 sampling test sets and was further validated in TCGA and METABRIC BC data. Our integrated study identified a 12-gene prognostic signature that could guide adjuvant therapy for BC patients and includes novel potential molecular targets for therapy.

Creative Commons License All site content, except where otherwise noted, is licensed under a Creative Commons Attribution 3.0 License.
PII: 14776