WASHINGTON, D.C., July 19 - Francis Collins on Wednesday opened a two-day meeting to discuss developing a public haplotype map of the human genome by declaring the event "one of the most important meetings we have organized in the last couple of years."
The approximately 150 participants, mostly academic scientists, with sprinklings of ethicists and industry executives, seemed equally enthusiastic about the task, though there was passionate debate about what populations should be included in the mapping project, and whether phenotypic identifying information should be collected along with the genotypes, among other issues.
What seemed of universal agreement was that any data generated in the proposed project should be free and publicly available.
The meeting was sponsored by the National Human Genome Research Institute, which Collins directs, though at times it was dominated by Eric Lander, the director of the Whitehead Center for Biomedical Research at MIT, who spoke during the introduction and moderated a panel discussion reviewing recent haplotype data. Both Collins and Lander stressed during the meeting that the goal of any haplotype map would be to design a set of tools and databases to link certain genes with diseases.
Because of the preliminary nature of the mapping discussions, Lander defined the scope of the enterprise early in the meeting. The broad goal, he said, should be to create "a sufficiently dense collection of SNPs genotyped in an appropriate collection to define haplotype blocks, common haplotypes for each block, and correlations between adjacent blocks in the human population."
Scientists who attended the meeting argued over a number of issues, including what populations to study, whether the samples should be families or unrelated individuals, how the density of the markers would affect their approach to locating SNPs, whether to map rare or common haplotypes, which analytical methods are appropriate for the task, and whether phenotypic identifiers should be included with the data.
The question of mapping certain populations was a particularly hot issue, veering across scientific and political discussions. Researchers disagreed on whether to study a single population or multiple populations, whether to do so only in the US or internationally, and how in fact to identify a population--on the basis of geography, ethnicity, language, or other criteria.
The debate did not produce definitive answers to these questions, but the group agreed to resolve them in future discussions.
Discussion grew passionate about genotyping Native Americans and Mexican Americans. Judy Gobert, Dean of the College of Math and Science at Salish Kootenai College in Pablo, Montana, said the project would disrespect Native Americans and pledged to fight the project. "I will go to every tribal government and tell them not to do it," she said. Her statement underscored that political and socioeconomic issues are intertwined with the scientific questions proposed for the haplotype mapping project.
The voice of industry was noticeably lacking at the meeting. Of the approximately 150 participants, only about 11 represented genomics companies.
"Few companies are doing haplotype mapping because many companies develop tools and not applications," explained meeting participant Michael Boyce-Jacino, chief technology officer at Orchid BioSciences. "I think as this evolves as a public data set, some companies that were tool providers will become applications players."
Jacino said that Orchid, which genotypes up to 700,000 to one million SNPs per week on a contract basis and for internal projects, views haplotype mapping as "a natural extension" of the company's current work.
The company arguably most deeply engaged in haplotype mapping, Genaissance Pharmaceuticals, presented data about haplotype variation and linkage disequilibrium in addition to participating in the discussions. Even so, Genaissance executive director of population genomics J. Claiborne Stephens said he would have liked to have been involved in the public haplotype discussion earlier on.
The current meeting "is a little more mature than a first meeting. It's disappointing we weren't called in before this," said Stephens.
But Stephens said the meeting validated Genaissance's four-year approach to mapping. "We turned the question around to not doing a full map of the whole genome but targeting certain pharmaceutically relevant genes," he said, adding that a public effort could be "complementary" to Genaissance's work.
"I'd like to see us involved [in the proposed public project]," he said. "We have the experience. If they're going to spend all this time, it seems reasonable that they should dovetail with what we've been doing."
Other companies with representatives who attended the meeting included Perlegen Sciences, Illumina, GlaxoSmithKline, Pharmacia, and Bristol-Myers Squibb. Illumina is developing large-scale SNP genotyping technology with Applied Biosystems, according to Jorge Velarde, Illumina's director of business development, and GlaxoSmithKline is also considering how to participate in a haplotype mapping project.
GlaxoSmithKline representatives were there "fact finding," said Clive Bowman, worldwide director of population genetics at GlaxoSmithKline, but "we are extremely interested in this approach." He added that the company was considering different approaches to haplotype mapping, including a large internal project, partnering with a biotech company, or working with the proposed public effort.
Ultimately, however, the success of the haplotyping effort hinges on drawing public support behind the project. At the close of the meeting, Collins said that finding a name for the haplotyping project that the layman could understand would be difficult, but necessary. "If we're going to capture the public's imagination we need some way this get this complicated genomics idea into common language," he said.