作者: Catherine A Ball , Alvis Brazma , Helen Causton , Steve Chervitz , Ron Edgar
DOI: 10.1371/JOURNAL.PBIO.0020317
关键词:
摘要: A fundamental principle guiding the publication of scientific results is that data supporting any scholarly work must be made fully available to research community, in a form allows basic conclusions evaluated independently. In context molecular biology, this has typically meant authors paper describing newly sequenced genome, gene, or protein deposit primary permanent, public repository, such as sequence databases maintained by DNA Data Bank Japan (DDBJ), European Bioinformatics Institute (EBI), and National Center for Biotechnology Information (NCBI). Similarly, we, members Microarray Gene Expression Society (MGED; http://www.mged.org), believe all journals should now require submission microarray repositories part process publication. While some have already condition acceptance, we feel requirements applied consistently recognize ArrayExpress (Brazma et. al. 2003), Omnibus (GEO) (Edgar al 2002), Biology Database (CIBEX) (Ikeo 2003) acceptable repositories. To end, MGED propose following new paradigm microarray-based studies. (1) Authors continue take responsibility ensuring collected analyzed their experiments adhere “Minimum about Experiment” (MIAME) guidelines use MIAME checklist (www.mged.org/Workgroups/MIAME/miame_checklist.html) means achieving goal. (2) Scientific are submitted one repositories—ArrayExpress, GEO, CIBEX—in format complies with guidelines. (3) Public establish release protocols assure compliance (4) To assist review process, collaboration publishers provide qualified referees secure accessing prepublication data. strongly encouraged submit during review. Naturally, protected from general prior either authorization submitters, whichever comes first. At minimum, valid accession numbers requirement publication, these included text manuscript allow community find access underlying data. Since its inception 1999, been working broader standards exchange annotation December 2001, proposed et 2001) requested interested parties feedback on relevance utility. The both researchers was overwhelmingly positive, yet almost everyone who responded also asked help implementing guidelines. Subsequently, summer 2002, an open letter various (e.g., Ball 2002a, 2002b) urging adopt We provided so could ensure sufficient information re-analyzed others would available. Again, response extremely most major publications comply standards. adoption greatly improved accessibility data, much it remains individual authors' websites variety formats; consequently, obtaining comparing datasets significant challenge. Clearly need additional include expression repositories. Though might ask why not original recommendation, answer quite simple—MIAME ahead time. NCBI EBI had developed nascent repositories, underway create similar database at DDBJ, submitting considerable burden authors. However, since time, improvements data-entry utilities GEO (www.ncbi.nlm.nih.gov/geo), (www.ebi.ac.uk/arrayexpress), CIBEX (cibex.nig.ac.jp), well growing number commercial academic software packages capable writing MAGE-ML documents (Spellman 2002) can directly databases, lowered barriers point where reconsider requirement. Requiring will distinct advantages entire community. These established commitment continued service providing level assurance published gene into future. Having standardized only make them more accessible, but integrated other relevant including genome sequences, single nucleotide polymorphism haplotype mapping information, literature, resources aid further interpretation patterns. Although many likely links current. Curation authors, reviewers, assuring requirements, enhancing standardization formats enable development analysis integration tools makes easier scientists access, query, share (5) Finally, confidentially, facilitating process. In same way availability profound impact wide range disciplines, requiring deposited necessity accelerate rate discovery. What proposal requires change which approach Both requisite available, because MIAME-compliant time effort, factored timelines. while may consuming painful first, benefits building repository far outweigh initial disadvantages. As always, our sincere hope suggestions stimulate discussion within together arrive consensus ensures widely easily accessible. Finally like urge EBI, towards exchanging