Data sharing: how much doesn't get submitted to GenBank?

A recent surge of interest in data sharing and data access has swept through the scientifi c community (e.g., [1]).Scientists recognize that free access to data is synergistic for fostering major advances.Concerns about standards of sharing are particularly acute with respect to large-scale DNA sequence and microarray data.Although some types of data have shallow histories or unclear protocols for how one would share them, DNA sequences have been deposited to the joint databases of GenBank, EMBL, and the DNA Databank of Japan [2,3,4] for over a decade, and many journals have policies requiring such submission before a paper can be accepted.For simplicity, we refer to these databases jointly as "GenBank."However, policies are only as good as their adherence and enforcement, and one should consider the effectiveness of past policies before launching new ones.Given its long history, an ideal case study is the deposit of sequences reported in published work into GenBank.We know from personal experience that authors of published papers reporting DNA sequences sometimes intentionally fail to deposit their sequences to GenBank and refuse to release them upon request.Is this a rare exception, or do many papers make it past coauthors, associate editors, editors, reviewers, and journal staff without providing the purportedly required data accession numbers?We examined the frequency with which published studies failed to submit their DNA sequences to GenBank.We searched six journals that have explicit policies requiring the submission of DNA sequences (Table 1).The previous six months (through February, 2006) of issues or most recent 30 papers presenting DNA sequence data were sought, though we examined a year of issues of one journal to test a longerterm trend (see below).Studies presenting only confi rmatory sequence or sequences that would not be found in nature (e.g., transgenes) were not counted.We searched for GenBank entries using locus and species names, authors of the article, etc., thus executing a search that would mimic any scientist's approach.No journal had complete compliance with its requirement for all DNA sequences to have been submitted to GenBank

Abstract licence: CC BY

5 more references, whose wording the paper’s licence keeps private.

• Public Data Archiving in Ecology and Evolution: How Well Are We Doing? · 2015

• What Drives Academic Data Sharing? · 2015

A person counts once a day per paper. We keep no cookie and no address: a reader is a hash that changes every day. Any month under five reads shows no number.