In a major restructuring of its Human Genome Program, the DOE Office of Health and Environmental Research announced the creation of the Joint Genome Institute to integrate work now based at its three major human genome centers. The merger represents a shift toward large-scale sequencing via intensified collaborations for more effective use of unique expertise and resources at Lawrence Livermore National Laboratory (LLNL), Lawrence Berkeley National Laboratory (LBNL), and Los Alamos National Laboratory (LANL). The genome centers are directed by Anthony Carrano (LLNL), Michael Palazzolo (LBNL), and Robert Moyzis (LANL).

"The Joint Genome Institute will enable more efficient cooperation among DOE laboratories to develop the next generation of genome sequencing technologies," said Martha Krebs, director of DOE's Office of Energy Research.

Elbert Branscomb (LLNL), newly appointed scientific director of the DOE Human Genome Program, also directs the joint institute, which begins operations in January 1997. Commenting on the rationale for the merger, Branscomb noted, "We need to focus our resources and integrate our efforts toward making a significant contribution to worldwide sequencing."

"Our goal is to commit an appropriate portion of our genome budget to sequencing a respectable fraction of the human genome with an efficiency at least comparable to that of other labs worldwide," he said.

In addition to production sequencing, a major goal of the joint center is to enrich the sequence data with information about its biological function, dubbed "functional genomics" as compared to the "structural genomics" of generating maps and other resources in the Human Genome Project. Functional genomics projects will focus on studying domains sequenced by the institute.

As the joint institute, the three laboratories will work together to define major tasks, each of which will be overseen by a manager. "This is a very high level of integration and merger," Branscomb said. "The managers and the joint institute will work together to make the best use of collective resources to achieve particular tasks." Palazzolo of LBNL will manage the genomic sequencing component.

The joint institute's budget will be task based, with allocations determined by external reviews of the proposed activities and contributions of each laboratory. Initially, the bulk of the funds will go toward large-scale sequencing, including production and technology development. The remainder will be used to develop and apply technologies for functional genomics.

Sequencing Targets, Quality, and Data Release

Gene-rich regions of around 1 to 10 Mb will be targeted for sequencing. Considerations include gene density, gene families (especially clustered families), correlations to model organism results, technical capabilities, and relevance to the DOE mission (e.g., DNA repair, cancer susceptibility, and impact of genotoxins). Use of mapped markers and Web sites will help avoid duplications in work. Targeting high-payoff regions will make data available to the broader research community.

Although the joint institute will strive for very high sequencing accuracy (around 99.99%), Branscomb pointed out that "there is much more to accuracy than just a number. You may get all the bases just right, but some clones may have been rearranged or aren't really from human DNA or may have other problems. The integrity of the sequence you generate is much more difficult to measure and presents a substantial challenge to all groups involved in production-scale work."

Data release will be immediate, comparable to procedures established by such major production-sequencing centers as Washington University, MIT-Whitehead, and Sanger Centre. Results will be posted daily on the Web; as data progresses to finished quality, it will be submitted to public databases.