Sequence data: posted vs. published.

Lee Rowen and colleagues highlight in their Policy Forum (15 Sept., p. 1881) their concerns with the publication of articles containing information derived from publicly available sequence data that have not yet been published in peer­reviewed primary papers. They refer to two review articles (1, 2) on the murine major histocompatibility complex (MHC) that were published before any primary papers from the data producers. In both cases, great care was taken to ensure that the source of primary sequence data was acknowledged in accordance with the guidelines set out by the hosting databases, the labs that produced the primary sequence were credited, and the European Molecular Biology Laboratory/GenBank accession numbers were quoted. It should be noted that Rowen's lab was one of those involved in the primary sequencing efforts, and the decision to publish was taken after a period of consultation with all those involved, including an offer of coauthorship. Although rejected, this latter offer is in line with the " agreeable options " for etiquette­led protocols on use of third­party database entries or annotations. At present, there is no worldwide policy governing the use of sequence data publicly available on genome databases. A major factor that was taken into consideration in the cases referred to above was the length of time that the primary sequence had been available in the public domain. We recognize that the value of work done by labs that contribute to publicly available genome data goes far beyond their skill in identifying and sequencing genes, and that they deserve priority in bringing the wider impact of these analyses to the primary literature. For the two MHC review papers referred to here, however, most of the primary sequences were deposited in the databases before the end of 1998. Sufficient time had therefore elapsed for the sequence producers to prepare, submit, and publish primary papers. We welcome the guidelines suggested by Rowen and colleagues in their Policy Forum. However, as they themselves admit, there will be instances where data producers do not publish their analyses within a reasonable time frame and that disputes will occur. They suggest that the research community develop a policy with guidelines about the kinds of analyses for which data producers can claim priority. Additionally, we would argue for a " statute of limitations " on such priority that would define a " reasonable time frame " in which sequence producers can publish …