easyDAS: Automatic creation of DAS servers

Gel Moreno et al. BMC Bioinformatics 2011, 12:23 http://www.biomedcentral.com/1471-2105/12/23 SOFTWARE Open Access easyDAS: Automatic creation of DAS servers Bernat Gel Moreno1,2*, Andrew M Jenkinson2, Rafael C Jimenez2, Xavier Messeguer Peypoch1, Henning Hermjakob2 Abstract Background: The Distributed Annotation System (DAS) has proven to be a successful way to publish and share biological data. Although there are more than 750 active registered servers from around 50 organizations, setting up a DAS server comprises a fair amount of work, making it difficult for many research groups to share their biological annotations. Given the clear advantage that the generalized sharing of relevant biological data is for the research community it would be desirable to facilitate the sharing process. Results: Here we present easyDAS, a web-based system enabling anyone to publish biological annotations with just some clicks. The system, available at http://www.ebi.ac.uk/panda-srv/easydas is capable of reading different standard data file formats, process the data and create a new publicly available DAS source in a completely automated way. The created sources are hosted on the EBI systems and can take advantage of its high storage capacity and network connection, freeing the data provider from any network management work. easyDAS is an open source project under the GNU LGPL license. Conclusions: easyDAS is an automated DAS source creation system which can help many researchers in sharing their biological data, potentially increasing the amount of relevant biological data available to the scientific community. Background In recent years the amount of biological data generated have been increasing greatly, and with the advent of new technologies like next generation sequencing this trend is likely to increase. Additionally, new analysis and re-analysis techniques help to produce better and more accurate derived results every day. Making all this data and results publicly available can be of great benefit for the scientific community as a whole, since valid biological data can be used both by field researchers and by those developing new methodologies and algorithms. Sharing raw research data allows others to conduct reanalysis and meta-analysis, usually reinforcing the previous results or even producing novel ones. ArrayExpress [1] and GEO [2] have had a very positive effect on microarray research development and have been heavily used in the development of both new data analysis techniques and biological knowledge. Sharing research results eases rapid spreading of new findings and its * Correspondence: 1 Software Department, UPC-BarcelonaTech, Barcelona, Spain Full list of author information is available at the end of the article incorporation in ongoing research increasing its overall usefulness. The effects of this sharing can be greatly increased if data and results are made publicly available using some kind of machine readable standard format allowing them to be seamlessly used by other researchers. While making the raw data available as supplementary material attached to the publication of a paper can be useful and other researchers can certainly use it, its integration with data from other sources will still be difficult and will not help fully automatic approaches such as workflows. However, if this same data is made available in a standard machine readable format it can be easily integrated with data coming from other standard sources and automatically displayed and analyzed. The Distributed Annotation System (DAS) is a complete system for sharing annotations on biological sequences. It comprises a standard XML based file format, an accurate definition of the semantics of the data -based on the use of ontologies of biological terms-, and an HTTP based REST style protocol for sharing those annotations [3-5]. Figure 1 is an overview of the DAS system. Many Annotation Servers can provide annotations © 2011 Gel et al; licensee BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. Gel Moreno et al. BMC Bioinformatics 2011, 12:23 http://www.biomedcentral.com/1471-2105/12/23 Page 2 of 8 Figure 1 Overview of DAS. Data is stored in either databases or data files (A). DAS servers (B) offer a common interface to access that data, the DAS protocol. Users (C), can use different client types (D) including specific DAS clients, internet browsers and non-visual scripts, to access the data. Optionally, clients can access the DAS Registry (E) to retrieve a list of available DAS sources. of sequence objects provided by a Reference Server, for example Ensembl or UniProt. While initially designed to annotate genomic sequences, DAS has support for both genomic and protein annotations, for sequence alignment data, and for structural information. Other federated systems exist for other types of biological data, such as PSICQUIC [6] for molecular interaction data. DAS client-server architecture was designed around the idea of having a small number of complex clients integrating data coming from, potentially, many of different simple sources. Some examples of DAS clients are Ensembl [7], Dasty2 [8], GBrowse [9], Jalview [10], SPICE [11], PeppeR [12], DASher [13]. Sharing biological data on DAS allows data providers to leverage the DAS ecosystem and make it easy to integrate their data with other existing sources. DAS server software is available in different programming languages, such as ProServer [14] in Perl and Dazzle [15] and MyDAS [16] in Java. However, despite the idea of DAS servers being simple, setting up a DAS server is not a trivial task. DAS servers allow for a great flexibility on where the actual data is stored and how it is structured. Usually the backend is database, but files and other options are also viable. The downside of this flexibility is that very often data providers will need to implement a custom made data access layer mapping their real data layout to the DAS concepts used in the server and this will have to be done either in Perl or Java. There are many research groups who will not have easy access to people proficient enough in programming to implement that access layer. In addition, setting up and managing an internet accessible machine to host Gel Moreno et al. BMC Bioinformatics 2011, 12:23 http://www.biomedcentral.com/1471-2105/12/23 the server can be also difficult or a too big overhead for many data generators, mainly for those with small data sets. Thus, the challenge: converting all those data generators into data providers, increasing the amount and variety of the biological data available to the scientific community and contributing to the collective annotation of biological sequences. Results and Discus (...truncated)