Given the exorbitant cost of deciphering genomes, most labs have given up sequencing and left that job to the big sequencing centers. But now, two groups have published methods that may make sequencing much cheaper and faster, promising small labs a chance to do it on their own.
Researchers all over the world still depend on a sequencing method introduced decades ago by Fred Sanger of the Laboratory of Molecular Biology in Cambridge, U.K. It uses bacteria to amplify the DNA and expensive reagents to label bases for identification. The cost has dropped since the mid-1990s from more than $1 to less than a 10th of a cent per base. But it's still high for many projects, including medical uses such as checking the genomes of individuals.
The new methods--one developed by George Church, a computational biochemical engineer at Harvard Medical School in Boston, and colleagues, and the other by Jonathan Rothberg, founder of 454 Life Sciences Corp. in Branford, Connecticut--save money by eliminating the need for bacteria and miniaturizing the process. In lieu of bacteria, DNA is attached to aqueous beads encased in oil, where chemical reactions do the copying. That change alone could reduce by two-thirds the costs associated with space and personnel, says Edward Rubin, director of the U.S. Department of Energy Joint Genome Institute in Walnut Creek, California. Moreover, both methods perform many thousands of these sequencing reactions at once in miniature "reactors," further bringing down the cost.
Once the DNA is ready, the two technologies diverge: The 454 technique puts the beads on a fiber-optic chip and uses flashes of white light to identify the bases. Rothberg washes the chip surface with one base at a time, creating four light patterns that a computer puts together as a sequence. He describes 454's success in sequencing Mycoplasma genitalium online 31 July in Nature.
Church's technique employs a microscope and other off-the-shelf equipment that use bursts of different fluorescent colors to distinguish the bases. Via this method, his team sequenced a strain of Escherichia coli and was able to detect easy-to-miss, single-base-pair changes from an almost identical E. coli genome. Church's group reports its results online today in Science.
Neither method is up to speed yet. The accuracy of both "should be improved by at least one order of magnitude," says Mostafa Ronaghi, a biochemist at Stanford University in Palo Alto, California. Also, to sequence mammalian genomes, the length of sequence generated, the "read," should be about 700 bases, but reads with these new approaches are hovering between 26 and 110 bases.
Whatever their limitations, the two reports signal the dawn of a new era in genome sequencing and detecting changes in individual genomes. Last year, the U.S. National Human Genome Research Institute in Bethesda, Maryland, began a program aimed at decreasing the cost of sequencing mammalian genomes to $100,000 in 5 years and to $1000 5 years later. That's what many think it will take for sequencing to become affordable in small labs.