BioEd Online

When Jonathan Rothberg's son was born six years ago, the baby was sent to the infant intensive care unit. Rothberg worried all night that something might be wrong with his child, and he found himself wishing he could just read the boy's genome to find out.

At the time that was impossible: it cost tens of millions of dollars and took more than a decade to decipher the first complete human genome, published in 2001 in Nature¹. But Rothberg's parental panic and frustration inspired him to design a faster, cheaper sequencing technique. Now Rothberg and his co-workers at the 454 Life Sciences Corporation, which he founded, report their success.

Writing in Nature², Marcel Margulies, Michael Egholm and colleagues describe a method that they say reads genomes 100 times faster than the current technology, which is based on a tried and true technique called the Sanger method.

Machines based on the Sanger method typically read 67,000 letters of the DNA code, also known as bases, in an hour; Rothberg says his method can decipher more than 6 million bases in the same time.

"We're a hundred times faster than Sanger, and we're just getting started," says Rothberg.

The 454 method is quick thanks to automation: the entire process from the initial multiplication of segments of DNA through to their sequencing is done using microfluidic technologies. It also analyses thousands of DNA molecules simultaneously. In contrast, Sanger sequencing takes many steps, and technicians are required to move the DNA from one stage to the next (see 'Cutting corners').

Using the 454 technique, one person using one machine could easily sequence the 3 billion base pairs in the human genome in a hundred days, Rothberg says.

To each their own

As the process gets faster, it gets less expensive. "It's clear that we'll be able to do this much cheaper," predicts Rothberg, who says that in the next few years scientists will be able to assemble a human genome for US$10,000.

That could bring about the long-touted 'era of personalized medicine', where drugs are tailored to an individual's DNA.

Several sequencing centres have already bought machines made by the 454 corporation, which is based in Branford, Connecticut. The technology was used to sequence the genome of the adenovirus in a single day, they reported in 2003. It was also used to sequence the bacterium that causes tuberculosis in a recent push to discover drugs against that disease.

The company has even offered to sequence James Watson, one of the discoverers of the helical structure of DNA. Watson says he gave some of his blood to the company this spring.

Error prone

Richard Gibbs, head of the sequencing center at Baylor College of Medicine in Houston, Texas, and a member of the 454 advisory board, says that the machines avoid some of the pitfalls of a bacterial cloning process that is part of the Sanger method. Certain pieces of DNA don't grow well in bacterial colonies, for example.

But the 454 machines are prone to their own sorts of errors, says Elaine Mardis, co-director of the Genome Sequencing Center at the Washington University Medical School in St. Louis.

The machines have trouble accurately reading long repeats of single base pairs, for example. Mardis says the machines are best used to analyse short genomes or short sequences. "At least in the short term, they won't replace the existing instruments."

"But they will provide us with very significant additional capacity to do specific kinds of projects," concedes Mardis.