Is community annotation the solution to constantly update databases?

In the last 8 years, extracellular vesicles (EVs) have attracted significant interest among scientists for their proposed role in intercellular communication, as reservoirs for disease biomarkers, and as targeted drug delivery vehicles. Multiple groups have reported the secretion of EVs and characterised their transcriptomic, proteomic and lipidomic content. In order to compare the data generated with other studies, researchers used to perform the cumbersome task of compiling the data from published EV studies. Since the launch of ExoCarta (https://www.exocarta.org) (Mathivanan and Simpson, 2009), a manually curated exosome-centric database that catalogs RNA, proteins and lipids, the process of comparison between datasets has been made easier.

However, growing interest in EVs has made updating and maintaining an online database a daunting task. Surveys of online links in the biomedical literature report that over 20% of database links are not active after their initial publication (Wren, 2004; Wren, 2008; Ducut et al, 2008). In addition, more than 50% of databases are never updated after initial publication, limiting their usability (Wren, 2008). The underlying reasons for this database decay are often lack of personnel and continued funding. A self-sustaining system that allows for users and researchers to contribute and update the databases may be a long term solution to the problem.

This daunting task of updating databases regularly and the need for a compendium for all types of EVs prompted the development of Vesiclepedia (https://www.microvesicles.org) in 2012 (Kalra et al, 2012). The Vesiclepedia compendium allows for community annotation. Since its initial release, through the active participation of the EV research community, the amount of data contained in Vesiclepedia has doubled. In addition to hosting mammalian data, Vesiclepedia also now hosts data from all organisms including prokaryotes. Furthermore, another database driven by community annotation, EVpedia (https://www.evpedia.info) (Kim et al, 2014), which catalogues EV data from prokaryotes and eukaryotes, was also recently initiated.

While semi-automatic measures are in place to allow for community annotation, the onus is on the research community in general to drive this effort forward. Researchers should develop a culture of depositing datasets to online resources to increase the visibility of the data. They should also develop the habit of participating in community annotations. Peer-reviewed journals also need to mandate the deposition of datasets to an online resource prior to publication (some journals indeed do this already).