The great accomplishment of biotech powerhouse Celera 鈥� the publication of an independent draft of the human genome 鈥� never really happened, claims a new report by members of the publicly-funded Human Genome Project.
They say new analysis reveals the company relied so heavily on publicly-available data that Celera鈥檚 accomplishment cannot be considered independent, or a victory for their 鈥渨hole genome shotgun鈥� (WGS) approach to genome sequencing.
鈥淭his has important implications for how we should sequence other large genomes,鈥� says Robert Waterston of the Genome Sequencing Center at Washington University in St. Louis, Missouri, one of the report鈥檚 authors. But Celera officials say the new study is seriously flawed.
Advertisement
The controversy is only the latest volley in a longstanding feud between Celera and the government and academic consortium over the sequence of our DNA. In 2001, even as both sides published genome drafts, they bickered over the benefits of their two approaches.
Theoretically faster
The public effort relied on a strategy where the genome was first broken into large, overlapping pieces, providing a low-resolution map of each chromosome. After the DNA sequence of each chunk was determined, the fragments could then be easily assembled into a complete sequence.
But Celera championed the WGS approach, in which all the DNA is shattered into random pieces. Once these are sequenced, powerful computers crunch the data to reassemble that jigsaw into a picture of the genome. Because it eliminates the need for meticulous isolation of each chromosome region, the shotgun approach is theoretically far faster.
The approach worked for smaller microbial genomes, and seemed successful for relatively simple animal genomes, such as the bacteria and fly. But critics questioned whether it could tackle complex mammalian genomes. The human and mouse, for instance, contain lots of repetitive DNA, which complicates computer analysis of random pieces.
The simultaneous publication of the sequences by the private and public groups last year in the journals Science and Nature respectively suggested both approaches had their merits. But, according to Waterston, Celera did not really stick with their plan.
Waste of time
He and his colleagues analysed the Science paper and say Celera merged the private and public data set in a way that retained the information of which sequences were contained in the same chunk 鈥� information Celera claimed it was a waste of time to gather. 鈥淚t gave them an easy path to reassemble each chunk,鈥� says Waterston.
Celera is not taking the criticism lying down. According to Gene Myers, Celera鈥檚 vice president of informatics research, Waterston鈥檚 study is seriously flawed because it only simulated the reconstruction of one chromosome. Myers says Celera鈥檚 success did not at all depend on the inclusion of the human genome project data. In fact, Myers鈥檚 team has an unpublished study, in which they re-ran the entire genome sequence assembly without the public data.
鈥淚n our paper, we included that data because we assumed more data was better,鈥� he says. 鈥淏ut the public sequence had so many errors in it, we got a better sequence by leaving it out. It was really surprising.鈥�
The sparks will probably continue to fly, since both teams are now working on the mouse genome. Waterson鈥檚 study, which is now available online, will be published with Celera鈥檚 rebuttal in a future issue of Proceedings of the National Academy of the Sciences.
Journal reference: Proceedings of the National Academy of Sciences (DOI: 10.1073/pnas.042692499)