The European Bioinformatics Institute in 2020: building a global infrastructure of interconnected data resources for the life sciences

Abstract

Data resources at the European Bioinformatics Institute (EMBL-EBI, https://www.ebi.ac.uk/) archive, organize and provide added-value analysis of research data produced around the world. This year's update for EMBL-EBI focuses on data exchanges among resources, both within the institute and with a wider global infrastructure. Within EMBL-EBI, data resources exchange data through a rich network of data flows mediated by automated systems. This network ensures that users are served with as much information as possible from any search and any starting point within EMBL-EBI’s websites. EMBL-EBI data resources also exchange data with hundreds of other data resources worldwide and collectively are a key component of a global infrastructure of interconnected life sciences data resources. We also describe the BioImage Archive, a deposition database for raw images derived from primary research that will supply data for future knowledgebases that will add value through curation of primary image data. We also report a new release of the PRIDE database with an improved technical infrastructure, a new API, a new webpage, and improved data exchange with UniProt and Expression Atlas. Training is a core mission of EMBL-EBI and in 2018 our training team served more users, both in-person and through web-based programmes, than ever before.

The international nucleotide sequence database collaboration

“…Since recent quantitative data rely on next-generation sequencers, expression information with metadata is registered in the functional genomics databases, Genomic Expression Archive (GEA) ( 2 ), Expression Atlas ( 8 ) and Gene Expression Omnibus (GEO) ( 9 ), while their raw sequence data are deposited in INSDC. To comply with informed consent agreements from donors, human genomes may be subject to access restrictions and are captured in sister databases operated by each partner that do not form part of INSDC: JGA at DDBJ ( 2 ), EGA at EMBL-EBI ( 10 ) and dbGaP at NCBI ( 11 ). Nonetheless, where appropriate, these databases share technical infrastructure and data standards with those of INSDC allowing interoperability.…”

From ArrayExpress to BioStudies

“…This will be addressed by supporting the ELIXIR AAI system ( 18 ), which will allow for a single login to all EMBL-EBI data resources. EMBL-EBI maintains a range of databases for specific types of life science data and knowledge ( 19 ), most with custom-built data deposition. The submissions of multimodal experiments to the EMBL-EBI resources is a more general question, that goes beyond the remit of this ArrayExpress update and here we discuss the relevant future developments only briefly.…”

Openness and trust in data-intensive science: the case of biocuration

“…Beyond the issue of how to deal with the “data deluge” lies the even more complex question of how to incorporate new research output with what is already known. As Attwood et al comment, new information is useless unless it is “stored and organized in ways that allow us to access it, to analyse it, to annotate it and to relate it to other information” (Attwood et al 2009 , p. 318) and in addition to serving as archives for research data output, several biological databases are so-called added-value databases, or knowledgebases , which “build on archival resources by providing expert curation, annotation, reanalysis, and integration of archived experimental data” (Cook et al 2019 , p. D17). A large part of biocuration work therefore consists of reading papers and translating relevant information into computer-readable formats, thereby enabling computational integration with new research output (Howe et al 2008 ).…”