Mining gene expression microarrays for long non-coding RNAs

Microarrays have been widely used in gene expression research. The amount of gene expression data has exponentially increased during the past decade. A great deal of data is stored in public data repositories like Gene Expression Omnibus and ArrayExpress, which can be freely accessed and exploited. With the development of high throughput techniques such as next generation sequencing, the number of long non-coding RNAs (lncRNAs) is highly increasing, while the number of protein-coding RNAs remains stable. LncRNAs play regulatory roles at almost every stage of gene expression from epigenetic modification to regulation of post-transcriptional process. Some lncRNAs are aberrantly expressed in disease states like cancer, Alzheimer's disease, and coronary artery disease. However, the research on lncRNAs is limited by the paucity of specific analytical tools. Some lncRNAs have been found on commonly used gene expression microarrays. A computational and manual re-annotation pipeline was developed to identify lncRNAs in traditional gene expression microarrays. Using this pipeline, one can get insights into the pathophysiologic function of lncRNAs. The approach motivates mining public and private microarray gene expression datasets to investigate the function of lncRNAs in disease development and progression. Ultimately, this may lead to the discovery of novel therapeutic targets.