We introduce the marine databases; MarRef, MarDB and MarCat (https://mmp.sfb.uit.no/databases/), which are publicly available resources that promote marine research and innovation. These data resources, which have been implemented in the Marine Metagenomics Portal (MMP) (https://mmp.sfb.uit.no/), are collections of richly annotated and manually curated contextual (metadata) and sequence databases representing three tiers of accuracy. While MarRef is a database for completely sequenced marine prokaryotic genomes, which represent a marine prokaryote reference genome database, MarDB includes all incomplete sequenced prokaryotic genomes regardless level of completeness. The last database, MarCat, represents a gene (protein) catalog of uncultivable (and cultivable) marine genes and proteins derived from marine metagenomics samples. The first versions of MarRef and MarDB contain 612 and 3726 records, respectively. Each record is built up of 106 metadata fields including attributes for sampling, sequencing, assembly and annotation in addition to the organism and taxonomic information. Currently, MarCat contains 1227 records with 55 metadata fields. Ontologies and controlled vocabularies are used in the contextual databases to enhance consistency. The user-friendly web interface lets the visitors browse, filter and search in the contextual databases and perform BLAST searches against the corresponding sequence databases. All contextual and sequence databases are freely accessible and downloadable from https://s1.sfb.uit.no/public/mar/.

Motivation The correct localization of proteins in cell compartments is a key issue for their function. Particularly, mitochondrial proteins are physiologically active in different compartments and their aberrant localization contributes to the pathogenesis of human mitochondrial pathologies. Many computational methods exist to assign protein sequences to subcellular compartments such as nucleus, cytoplasm and organelles. However, a substantial lack of experimental evidence in public sequence databases hampered so far a finer grain discrimination, including also intra-organelle compartments. Results We describe DeepMito, a novel method for predicting protein sub-mitochondrial cellular localization. Taking advantage of powerful deep-learning approaches, such as convolutional neural networks, our method is able to achieve very high prediction performances when discriminating among four different mitochondrial compartments (matrix, outer, inner and intermembrane regions). The method is trained and tested in cross-validation on a newly generated, high-quality dataset comprising 424 mitochondrial proteins with experimental evidence for sub-organelle localizations. We benchmark DeepMito towards the only one recent approach developed for the same task. Results indicate that DeepMito performances are superior. Finally, genomic-scale prediction on a highly-curated dataset of human mitochondrial proteins further confirms the effectiveness of our approach and suggests that DeepMito is a good candidate for genome-scale annotation of mitochondrial protein subcellular localization. Availability and implementation The DeepMito web server as well as all datasets used in this study are available at http://busca.biocomp.unibo.it/deepmito. A standalone version of DeepMito is available on DockerHub at https://hub.docker.com/r/bolognabiocomp/deepmito. DeepMito source code is available on GitHub at https://github.com/BolognaBiocomp/deepmito Supplementary information Supplementary data are available at Bioinformatics online.

ABSTRACT