Letters to the editor should be sent to:
Links to Other ASM Pages:
Microbial Genome Sequencing
Microbial Genomes: a Blueprint for Life [PDF]
The American Academy of Microbiology (AAM) recently released a report on microbial genome sequencing projects. This summary of a colloquium convened by the AAM 19-21 March 2000 in New Orleans, La., raises a number of important issues related to the recent developments in microbial genome sequencing. The conclusions encourage ASM to "play a proactive role in keeping the microbial genome sequencing endeavor vigorous and productive." The outline and overall direction of the report are laudable; however, before ASM accepts all the recommendations, I think the data release policies promoted by the authors deserve careful consideration.
MICROBIAL GENOME SEQUENCING PROJECTS
In several places the report proposes allowing researchers to withhold genomic sequence data from public purview. The conclusion that "data should be released either at the time of publication or when the grant that funded the sequencing expires" implies that sequence data may be justifiably withheld from public examination and use for essentially as long as the principle investigator desires. This advice contradicts the guidelines released from the NIH that microbial genomes should be released within one month of reaching threefold coverage. In practice, the majority of publicly funded genome sequencing projects (with a few notable exceptions) opt for immediate or near-immediate data release policies.
Suggesting that the experimenter withhold the sequence data until they have finished their analysis suggests that the only important question about that genome is the one that the sequencer is asking. Furthermore, involvement in a genome sequencing project has been sufficient scientific benefit to encourage many productive collaborations. Experience has already shown that a wide variety of questions can be answered, once a sequence is publicly available—more than a single researcher could answer in a lifetime. Furthermore, far from limiting the ability of the individual researcher to analyze the genome, the rapid and frequent dissemination of data results in discussions between interested parties that ultimately benefit all involved. Thus, the benefits of widespread dissemination outweigh the marginal costs.
In spite of NIH guidelines, some project leaders have chosen to retain all data from sequencing projects, or allow only BLAST searches which render novel portions of the data inaccessible. This results in the sequencers becoming divorced from both related projects and the biology behind their sequencing. Retaining sequencing information "until publication" also allows sequencers to favor some researchers. This provides an unfair advantage to "friends" of the researcher over other members of the scientific community, a much more egregious assault on scientific integrity and ethics than using public sequence data with appropriate attribution. The biology must drive the sequencing and not vice versa.
TIGR
Microbial Genomes Blast DatabasesThe report also criticizes the lack of central organization of genomic sequencing, claiming it leads to redundancy. However, duplication of sequencing effort has been fostered by delaying sequence data release. Other groups that are considering sequencing related organisms and do not know the status of "private" programs may proceed unilaterally, resulting in duplicated effort. Many biologists would rather duplicate a sequencing project if it ensured public data release instead of waiting for the sequence to be released island-by-island or never at all. Similarly, the report calls for a central repository for genomic sequence and mentions the TIGR website as a leading resource. However, the report completely ignores the critical role of the NIH microbial genomes page in the dissemination of bacterial genome sequencing data. The primary function of this site is to provide a single point of access to data generated from the diverse microbial sequencing projects regardless of funding source. Secondary functions include informing others about current sequencing projects (to reduce redundancy) and providing links to information centers. Currently data release to this site is voluntary and incomplete. For example, many state- or privately funded projects are included, while some key NIH-funded projects are not included. Future funding should demand frequent data release via this publicly funded national repository.
The joint Clinton-Blair statement mentioned in the report states that [human] sequence data should be released "for the benefit of researchers in every corner of the globe." This policy is required for all projects funded by the National Human Genome Research Initiative and should be required of all publicly funded genomics projects. By promoting frequent and public data release, ASM will demonstrate appropriate public-spirited leadership. ASM needs to review the recommendations of the AAM report, and consider whether a closed data release policy is really an appropriate use of taxpayers' money given the potential benefits to society. The only way ASM can play a "proactive role in keeping the microbial genome sequencing endeavor vigorous and productive" is to promote an open and instant data release policy that benefits all researchers in every corner of the globe.
Robert Edwards
University of Tennessee Health Sciences Center, Memphis
redwards@utmem.eduP.S. The author is a collaborator on several Salmonella sequencing projects. More information on these projects and all the sequence data can be retrieved from salmonella.org.
Email: webmaster@asmusa.org