Criticized and rejected, the whole-genome sequencing shotgun approach sparked an often bitter battle between its main opponent, the publicly funded Human Genome Project, and its most visible champion, Craig Venter, Celera Genomics president and chief scientific officer.
In a recent about-face, public researchers conceded that the method they once scoffed at has its benefits and are now poised to incorporate it in their effort to sequence the mouse.
Given the remarkable similarity between the mouse and human genomes and the rodents prominence as a clinical model for human research, the concession is a noteworthy validation of the oft-disparaged approach if its good enough for the mouse, chances are its good enough for anything.
But the public effort is not completely doing away with the BAC-based method. Instead it plans to combine the two methods to exploit the benefits of each.
Early on we thought we had to do things BAC-by-BAC and Celera argued you could do it whole-genome shotgun, Eric Green, director of the National Institutes of Healths Intramural Sequencing Center said. What everybody now realizes is that the ideal way is to do a little bit of both.
Celera spokeswoman Heather Kowalski said, Its not surpising that NIH has adop-ted the shotgun approach. She noted that the chief of the Human Genome Project, Francis Collins, even went as far as to praise Celeras methods at the joint human genome sequencing announcement back in June.
In the whole-genome shotgun approach, researchers chop the entire genome into small, random, overlapping pieces, sequence them, and then reassemble the fragments to recreate the genome.
Venter first successfully tested the whole-genome shotgun method on the genome of the bacterium H. influenzae, but the Human Genome Project rejected the approach as unworkable.
There was no way at that point in time to take all those reads and assemble them to represent the genome, said John McPherson, head of the mouse sequencing project at the Washington University Genome Sequencing Center in St. Louis.
Given the numerous repeat-rich regions in the human genome, McPherson said, how would you know you got them assembled correctly?
The public project insisted on using a BAC-based strategy whereby researchers create bacterial clones containing overlapping 100 to 200 kilobase DNA fragments from known locations of the genome. Each BAC, as the clones are called, is then sequenced individually.
Venter, confident that human-genome shotgunning was not only practical, but also speedy and cheap, created Celera and gave the Human Genome Project a run for its money, driving human genome sequencing to completion four years ahead of schedule.
Eager to prove that shotgunning was practical for decoding a large genome, Celera collaborated with the Berkeley Drosophila Genome Project and in February the private-public team announced it had decoded the fruit fly.
Together they used a hybrid approachCelera provided a whole-genome shotgun sequence while Berkeley provided the BACs.
The public mouse project, inspired by the success of the Drosophila project, is now convinced that shotgunning is worth a shot.
Predominantly most of the mouse genome will be sequenced at two centersthe Whitehead Sequencing Center and the Wash. U. Sequencing Center, using a combined approach of a whole-genome shotgun combined with BAC-by-BAC sequencing, reminiscent of the way the Drosophila genome was sequenced, Green said.
Having tested the strategy on Drosophila it was recognized as being worth doing on the mouse, he added.
Each approach has its advantage. The whole-genome shotgun method has the advantage of generating huge amounts of data very quickly. But putting the puzzle back together is very complicated. Assembly is much simpler in the BAC-based approach, because each DNA fragment can be traced back to a specific BAC. But the process is much slower. By combining both approaches, Green said, you can place the whole-genome shotgun reads into a confined bin the size of a BAC.
In other words, the researchers will use BACs as a scaffold on which to drape the whole-genome shotgun sequence. To create the BAC scaffold of the mouse genome they will randomly select representative BACs or well-mapped BACs for low sampling across each one.
The public sequencing project will also use the same strategy when it tackles the rat genome.
The use of this hybrid approach means genomewide mouse data will be available in the early stages of the project providing researchers annotating the human genome with a powerful tool for comparisons.
In a sense, Green said, Celera has been using the combination approach all along. In the case of human what does Celera do? Celera does a whole-genome shotgun but then they can get off the Internet the public effort, which is BAC-by-BAC, he said. So they are actually doing a hybrid in their own way.
Its not a matter of one winning or the other winning, said Green. They said whole genome shotgun. The public effort said BAC-by-BAC. The answer is probably a hybrid.
Aaron J. Sender