Genomics: doing Moore with less

Genomics is experiencing its own manifestation of computing technology's Moore's law of cost and efficiency, says Prof Richard Gibbs: the volume of sequence data is increasing by around 10-fold every year as sequencing costs continue to plummet.

The Australian-born geneticist, director of Baylor College of Medicine's Human Genome Sequencing Centre (HGSC) in Houston, Texas, will be among the luminaries at the International Conference of Genetics (ICoG 2003) in Melbourne in July.

Gibbs, who completed a PhD in genetics and radiation biology in 1986, joined Baylor as a postdoctoral researcher to study X-linked genetic diseases in humans and began developing new technologies for rapid DNA sequencing.

He was appointed a faculty member in 1991, and when Baylor was selected as one of three original institutions to participate in the NIH-funded Human Genome Project five years later, became founding director of the new HGSC.

Gibbs says new DNA sequences are now pouring into international databases at a rate of 1.5 billion bases per month, or around 50 million bases a day. "We contributed about 10 per cent of the human genome over a period of many years, and did a third of Drosophila," Gibbs says.

"We've now completed half the rat genome in only two years, while our staff has fallen from 250 to 190 in the same period, reflecting the increasing efficiency of automated sequencing."

Since he went to Baylor, new automated, rapid-sequencing technology has driven down the cost-per-base sequenced from $US14 to just 6 cents. The next challenge for the centre is to try to complete a mammalian genome, on its own, in a year.

Since the Human Genome Project, the Baylor Centre has become involved in an eclectic but strategic tour of the gene pools of some of geneticists' favourite species. They include the rhesus macaque, rat, sea urchin, cow, European honeybee, the fruit fly Drosophila pseudobscura, a cousin of D. melanogaster, and, in Gibbs' words, "a bit of chimp" -- Boston's Whitehead Institute has carriage of the Pan troglodytes genome project.

These species have been chosen to provide maximum insight into the shared genetic heritage of distantly related species. Comparative analyses should illuminate the most significant genetic changes -- and by inference, the selection pressures -- that fuelled the evolutionary divergence of closely related species, like the rat and mouse, humans and chimps, and, most recently, Drosophila fruit flies.

Here, says Gibbs, comparative genomics makes it possible to look for hotspots -- chromosomal changes, or genes that have undergone significant modification relatively recently. These are likely to be most informative about why -- and when -- species diverged.

"We can already see some significant differences between the mouse and rat genomes, but they're somewhat artefactual because neither project is complete," he said.

"There are only 100-odd chromosomal rearrangements between them. When you align the corresponding segments, the gene order and gene sequences are still very similar -- the rat and mouse share about 70 per cent genetic identify, and each overlaps with human by around 40 per cent.

"As the third mammal species to be sequenced, the rat adds value by serving as an anchor for the other two."

The evolutionary hot spots in genomes tend to cluster in regions lying close to centromeres -- the knot-like attachment points for the spindle fibres that pull duplicated chromosomes apart during cell division.

Current estimates suggest chimps and humans differ by only 2 per cent of their DNA -- yet those differences underlie major changes in anatomy, brain volume and brain function in humans. The rhesus macaque, more distantly related to human and chimp, probably differs from either species by around 10 per cent, and will provide information about the divergence of the great ape and monkey lineages.

Gibbs says the differing frequencies of much more recent evolutionary innovations, so-called single-nucleotide polymorphisms (SNPs) should yield the finer detail of recent human migrations, their timing, and of genetic encounters between different races and cultures.

The next step will be large-scale genotyping of individuals within a species, or regional populations -- notably human. But the cost-per-base must fall even further.

Could such data yield some of the fine detail of recent human evolution, and human movements in prehistoric times?

"That's an interesting question," says Gibbs. "We're looking to plot these genetic data, and develop molecular clocks and markers for migration. Once you understand the population structure, from the frequency of these markers, we're going to see parts of chromosomes that will be common in certain areas of the world, but rare in others. From these, you can make inferences about migrations.

"It's good stuff, but it's also scary stuff because of how racists might use it."

But Gibbs says information from mass-genotyping should be invaluable in identifying SNPs associated with inherited genetic disorders or predispositions to genetically complex health problems such as cardiovascular disease.

"That's where it all started," he said. "It will be nice to complete the circle, for people wondering what all the hype was about."

Australian Biotechnology News is a major sponsor of the 19th Annual Congress of Genetics. This story is the latest in a regular series presented in the magazine in the lead-up to this event.