Exploratory meta-analysis of hypoxic transcriptomes using a precise transcript reference sequence set

Abstract

Gene expression studies are intrinsically biased, with many studies influenced by concomitant information such as gene–disease associations. This limitation can be overcome using a data-driven analysis approach without relying on ancillary information. The FANTOM CAGE–Associated Transcriptome project provides a comprehensive meta-assembly of the human transcriptome using coding and noncoding genes. Hypoxia strongly influences gene expression; in addition, noncoding RNA (ncRNA) metabolism is down-regulated in response to hypoxic stimuli. We evaluated the differential response of various transcripts to hypoxia by determining their hypoxia responsiveness scores. Enrichment analysis revealed that several genes associated with ncRNA metabolism, particularly those involved in ribosomal RNA processing, were down-regulated in response to hypoxia. Previously published information from the FANTOM CAGE–Associated Transcriptome project was suitable for meta-analysis of the transcriptome sequencing data from both coding and ncRNAs and to evaluate the hypoxia responsiveness of target transcripts and relationship between sense–antisense transcripts from the same locus. Our results may facilitate functional annotation of various transcripts including ncRNAs, allowing for both sense and antisense and coding and noncoding evaluations.