Facilitating Comparative Functional Genomics through Data Reuse with Livestock EpiDB

Livestock EpiDB was designed to facilitate data sharing, integration and mining of public next-generation sequencing (NGS) data. It is comprised of a web portal, database and data processing pipelines that identify, filter and process data from both NCBI’s SRA and EBI’s ENA databases. Filtering occurs in two steps where the first step identifies only well-annotated data for query through a MySQL database and the second captures only high-quality sequence for processing to use for the gene-expression atlas and meta-analyses. Data types captured within EpiDB includes: RNA-seq, miRNA-seq, methyl-seq, and ChIP-seq data. Over 12Tb of livestock data has been queried to date, yielding a subset containing more than 4000 well-annotated NGS samples across five livestock species (cattle, chicken, horse, pig, sheep). This data represented 106, 30, 16, 24 and 16 tissues per respective species (cattle, chicken, horse, pig, sheep). Metadata encompassing experimental details can be downloaded using the search tool to identify specific data by species, data type and tissue from the database. A gene expression atlas also provides standardized expression levels for both mRNA and miRNA across tissues. Alternatively, users can download normalized, pre-processed data or access this data through the CoGe genome browser for comparison to their own existing datasets. Users will soon be able to download allele counts from public data for use to identify allele specific expression. EpiDB data and resources are available for use online at http://www.animalgenome.org/epidb/index.php