No bull: Upholding community standards in public sharing of biological datasets

We write as representatives of a multi-institutional team working since 1997 to sequence the Y chromosomes of eight mammals (human, chimpanzee, rhesus macaque, marmoset, mouse, rat, bull, and opossum) to exacting standards of accuracy and completeness; we are pleased to say that this goal is in sight. Throughout the course of the project, we have publicly shared the immediate products of our efforts [the completed sequences of individual bacterial artificial chromosome (BAC) clones] via GenBank, in keeping with the 1996 Bermuda Principles (www.genome.gov/10506537), so the data could be used by the community. We have done so with the understanding that our research team’s interests are protected by the 2003 Fort Lauderdale Agreement (www.genome.gov/pages/research/wellcomereport0303.pdf) and the 2009 Toronto Agreement (www.nature.com/nature/journal/v461/n7261/full/461168a.html), which govern the use of publicly deposited biological datasets. Specifically, the Fort Lauderdale and Toronto agreements protect the right of data generators to publish the first analyses of their publicly deposited datasets. Taken together, the Bermuda, Fort Lauderdale, and Toronto accords have promoted and enabled the immediate public sharing of large and valuable biological datasets. The record shows that these widely accepted principles have served the interests of the public and of the scientific community as a whole.

It is in this context that we were surprised to read the publication by Chang et al. (1), which presents an unauthorized annotation of the entirety of our preliminary bull Y chromosome sequence assembly (GenBank accession no. CM001061). We first announced the plan to assemble and annotate the bull Y chromosome in 2006 (2). Professor Liu made unauthorized use of our research team’s unpublished data in publications in 2009 and 2011 (3–5), and we communicated our concerns to Professor Liu after reading these papers. We informed Professor Liu of our plans to publish a comprehensive and systematic description and analysis of the bull Y sequence on completion of the project, and we requested that he respect the Fort Lauderdale and Toronto guidelines. The recent publication (1) is decidedly at odds with these community principles, and as such poses a direct challenge to the practice of immediately and publicly sharing broadly useful datasets.

Beyond these concerns, we also write to raise issues regarding the scientific accuracy and validity of Chang et al.’s analyses and conclusions regarding the bull Y chromosome sequence. For example, we find that the majority of the 1,274 bull Y genomic sequences that Chang and colleagues identify as protein-coding genes are actually pseudogenes, many of which contain truncating mutations and many of which lack compelling evidence of transcriptional activity. It is of note here that the high-throughput mRNA sequencing datasets alluded to by Chang and colleagues have not been publicly released.

Our research team will in due course publish a rigorously grounded assembly, analysis, and annotation of the bull Y chromosome. In the meantime, our team will continue to publicly and immediately share the genomic sequence datasets that the team generates during the closing phases of this technically challenging project.