By Monica Heger
This article has been updated from a version posted June 8 to correct information provided by Biesecker regarding the use of Illumina's HiSeq 2000.
The National Human Genome Research Institute has made significant progress on its ClinSeq study, moving from targeted sequencing of candidate genes using capillary sequencing to whole-exome sequencing on the Illumina Genome Analyzer with Agilent's SureSelect, In Sequence has learned.
In addition, for two patients involved in the study, NHGRI researchers have sequenced the entire genome on the Illumina platform.
So far, the project has completed the sequencing of 250 candidate genes in about 700 patients. The researchers are now in the process of analyzing the finished sequence results and are continuing to sequence about 15 to 20 exomes each week.
Leslie Biesecker, project head and senior investigator of NHGRI's genetic diseases research branch, spoke about the project at a workshop for science writers held in Washington D.C. this week.
Biesecker said that even though the cost of whole-genome sequencing is continuing to drop, for now the effort will stick with exome sequencing because it is more cost-effective. The reagents cost for exome sequencing is around $2,500 per patient, while for whole-genome sequencing it is still between $10,000 and $15,000, he said.
The ClinSeq project is a large-scale medical sequencing project focused on uncovering the causative hereditary mutations in atherosclerosis (IS 12/23/2008). Since the project began in 2007, it has increased the number of patients it will recruit from 1,000 to 1,500 and has now moved into next-gen sequencing.
The goal is to "pinpoint risks and give individualized risks and tailored treatment to patients," said Biesecker. He said the project chose atherosclerosis because it is a complex disease influenced by genetic components, including both rare and common variants, as well as environmental and behavioral risks.
"We want to change the medical paradigm," Biesecker said. Instead of waiting for symptoms to develop, patients could be evaluated for their risk of heart disease beforehand.
Biesecker discussed one case in particular, a woman who did not have any symptoms of heart disease, but was recruited to the study because she had a family history. After sequencing the 250 candidate genes, the researchers discovered that she contained a rare variant associated with early-onset heart disease. "This is potentially a very powerful approach to medicine," he said.
Rick Del Sontro, a ClinSeq participant who spoke at the workshop, said that he was drawn to the study because he has a family history of heart disease at young ages, despite good eating and exercise habits. His mother died at age 69 of heart disease after undergoing her third bypass surgery, and his sister had two stents by age 47.
Now, Del Sontro and his seven siblings have all enrolled in the ClinSeq project. While the researchers have so far not been able to uncover the genetic cause of the family's heart disease, they have found other clinically relevant variants. For instance, they determined that Del Sontro had hereditary neuropathy pressure palsy, characterized by numbness in the extremities. He said that they also found other genetic diseases for which he was a carrier.
Biesecker said that making sense of all the data and translating variants into meaningful assessments of "phenotype and developmental predictions" has proven to be challenging. It's a "fire hose of data" he said.
Other researchers agreed that the major bottleneck of bringing sequencing to the clinic is in analyzing and interpreting the results. "It now costs more to store and analyze the sequence data than it does to do the sequencing," said Vivien Bonazzi, program director for computational biology and bioinformatics at NHGRI.
Moving forward, she said it would be important to develop new analysis tools, and actively engage the biological and computing communities to make interpretation more clear.
Another challenge of the ClinSeq project is that, unlike many other sequencing projects, the results are returned to the participants. Biesecker said one of the reasons for doing that was to try and figure out the best methods for delivering sequencing results. As sequencing costs continue to plummet, and direct-to-consumer sequencing becomes more widely available, it will be important to determine best practices in delivering results to the patients.
"We are trying to figure out the subsets of data that the patient will want, and how they will use it and interpret it," he said. For instance, they are trying to figure out whether patients would still want results even if the interpretation was unclear or not medically relevant.