A Turning Point for Personal Genomes
Scientists are finally starting to find medical information of value.
Last year, when more than 100 of the world’s top geneticists, technologists, and clinicians converged on Cold Spring Harbor Laboratory in New York for the first annual Personal-Genomes conference, the main focus was James Watson’s genome. The codiscoverer of the structure of DNA was the first to have his genome sequenced and published (aside from Craig Venter, who used his own DNA for the private arm of the human genome project.) Watson sat in the front row of the lecture hall as scientists presented their analysis of his genome. They paid special attention to the number of single-letter variations or small insertions and deletions in his DNA–clues as to whether he had a genetic variation that slightly boosted his risk for heart disease or cancer. But there was very little usable information in the genome.
That has all changed. In the last year, the number of sequenced, published genomes has shot up from two or three to approximately nine, with another 40 or so genomes sequenced but not yet published. “While the numbers are still small numbers, we are starting to put this research into the real disease context and get something out of it,” says Jay Shendure, a geneticist at the University of Washington in Seattle, and a TR35 winner in 2006.
Last year, sequencing a genome was still a feat in itself, and much of the conference focused on the technical details–assessing accuracy and error rates and comparing one method to another. While these issues are still of central importance, sequencing a human genome has become routine enough to generate medically useful information. “Now we are able to do things automatically, so the biology starts to come out,” says Paul Flicek, a bioinformaticist with the European Bioinformatics Institute and one of the conference organizers.
In a few cases, scientists have already been able to find the genetic cause of a disorder by sequencing an affected person’s genome. Shendure has sequenced the coding region–the 1 percent of the genome that directs production of proteins–of the genomes of a handful of families with children afflicted with a rare inherited disorder called Miller Syndrome, which is linked to facial and limb abnormalities. Researchers compiled a list of genetic variations in each person and filtered out those that have been commonly found in people without the disease variations. They then looked for variants present only in affected people, and came up with one candidate gene. Shendure declined to identify the gene prior to publishing the findings, but noted that it was one they would not have anticipated. He hopes the technique can be applied to more common diseases as well, perhaps by studying people with early onset or extreme cases.
Genome sequencing has also engendered a new approach to cancer research. Last year, Elaine Mardis and her team at Washington University School of Medicine in St. Louis sequenced the complete genomes of cancerous and normal tissue in a patient with acute myeloid leukemia, identifying 10 mutated genes that appear to play a role in this cancer. This year, her team has sequenced the genome of four different types of tissue from a breast-cancer patient–the normal genome, DNA from the primary tumor, DNA from a metastatic brain tumor (a secondary tumor formed from cancer cells originally from the breast tumor), and DNA from the patient’s cancerous tissue implanted into a mouse. (Because the cancerous tissue removed during surgery is often inadequate for genetic research, scientists sometimes grow tumor tissue from the patient’s cancer cells in mice.)
While the vast majority of the sequence will be identical in all four samples, identifying differences could pinpoint the genetic changes that lead to the initial formation of the tumor, as well as those that trigger metastasis. If scientists can find drugs that block the primary tumor from spreading, cancer could be converted into a manageable chronic disease.
Mardis’s team has already identified a number of variants that are unique to either the primary tumor or the metastatic tumor. They have also found some variants that appear in both but are more common in the metastatic tissue, suggesting that this type of mutation might enable cells to spread through the body. “We are now looking at breast-cancer-derived brain, lung, and liver tumors to see if there are commonalities in metastatic disease,” says Mardis. Her center aims to sequence 150 cancer genomes this year. Next year, that number will likely seem small.
Become an MIT Technology Review Insider for in-depth analysis and unparalleled perspective.Subscribe today