Posted on behalf of Chris Gunter
Believe it or not, some people are complaining that now the ASHG meeting has too much genomics. This comes after years of dire warnings that the society was ignoring genomics at its own peril. Friday morning featured yet another all-genomics platform session, with multiple talks on sequencing whole human genomes.
Matthew Bainbridge of Baylor made crowd laugh by saying all we need is an assay with 100% sensitivity and 100% specificity. In the meantime, he’s using a method of sequencing a genome on two different platforms and comparing the results. He and others at the Baylor College of Medicine extended their work on personal genomes by sequencing in a family with a form of ataxia not due to a triplet repeat mutation, as many ataxias are. As in several talks at the meeting, Bainbridge and colleagues captured the sequence of the exome (or entire set of expressed genes) from key members of a pedigree, and by comparison of genetic variants were able to pinpoint the one gene they think carries the responsible mutation.
In the same session, Stephan Schuster of Penn State discussed the genome sequence of four Kalahari bushmen. He claimed that these were the first genome from non-pastoralists, so they were likely to have different metabolism than other sequenced genomes. Two points stood out for me: first, he addressed the motivation on the part of the Kalahari for participating in the project, as they are fully consented for the full genome sequence data to be freely released. “Desired outcome for Bushmen: they hope that via this project their unique position among today’s human population will be documented and they will benefit thru inclusion in ongoing pharmacological studies.”
Second, Schuster showed a maximum likelihood evolutionary tree of the few whole human genomes currently available. The branch groupings were determined not by the actual evolutionary history but by the method of sequencing used to generate the data. Thus “we are comparing apples to oranges,” he said, “and need to come up with a new set of rules for whole-genome phylogenies.” The crowd agreed as there was much murmuring and furious typing. Schuster proposed a number of reasons why this would happen; solutions include standardizing SNP calling between platforms and agreeing on minima for sequencing coverage.
This brought to mind the presidential address from Ed McCabe earlier in the week – McCabe related his own experience with proprietary algorithms created by companies to analyze the data produced with their technologies. He said that one of his trainees had written an abstract for last year’s meeting and was about to send it off. As a double-check, she reran the analyses first, and got completely different results. The company had changed the algorithm without telling them. When they contacted the company, he and the trainee were told essentially “yes we know there are some problems with the new update but we figure once more people use it and know about the problems, some one out there will come up with a solution.” McCabe therefore encouraged ASHG members to become more involved with the society’s Corporate Responsibility Task Force, launched in 2009 to handle issues just like these.
More details are in the February 2009 issue of the ASHG newsletter.