Mammalian Annotation Database for improved annotation and functional classification of Omics datasets from less well-annotated organisms

Next-generation sequencing technologies and the availability of an increasing number of mammalian and other genomes allow gene expression studies, particularly RNA sequencing, in many non-model organisms. However, incomplete genome annotation and assignments of genes to functional annotation databases can lead to a substantial loss of information in downstream data analysis. To overcome this, we developed Mammalian Annotation Database tool (MAdb, https://madb.ethz.ch) to conveniently provide homologous gene information for selected mammalian species. The assignment between species is performed in three steps: (i) matching official gene symbols, (ii) using ortholog information contained in Ensembl Compara and (iii) pairwise BLAST comparisons of all transcripts. In addition, we developed a new tool (AnnOverlappeR) for the reliable assignment of the National Center for Biotechnology Information (NCBI) and Ensembl gene IDs. The gene lists translated to gene IDs of well-annotated species such as a human can be used for improved functional annotation with relevant tools based on Gene Ontology and molecular pathway information. We tested the MAdb on a published RNA-seq data set for the pig and showed clearly improved overrepresentation analysis results based on the assigned human homologous gene identifiers. Using the MAdb revealed a similar list of human homologous genes and functional annotation results regardless of whether starting with gene IDs from NCBI or Ensembl. The MAdb database is accessible via a web interface and a Galaxy application.