The cost of Sanger sequencing
, compared with available next-gen platforms, prompted Agencourt Bioscience to use its fleet of Applied Biosystems capillary-electrophoresis instruments to deliver sequencing services for several projects under its new contract with the National Cancer Institute’s Office of Cancer Genomics (see In Sequence 11/13/2007).
The one-year contract, which can be renewed for a second year, calls for Agencourt, a subsidiary of Beckman Coulter, to sequence genes and genomic regions from hundreds of cancer samples provided by multiple investigators. Its work will serve several NCI-funded projects, including the Childhood Cancer Therapeutically Applicable Research to Generate Effective Treatments initiative, according to the OCG.
Cost influenced Agencourt’s decision to use its approximately 35 ABI 3730xl sequencers. When the company was preparing its bid ahead of the March NCI deadline, it “reviewed the cost for Sanger vs. what was available at the time, which was 454, and we felt that the cost for Sanger was actually quite competitive in terms of meeting the high quality criteria,” Erick Suh, director of genomic services at Agencourt, told In Sequence last week.
“The cost of sequencing with Sanger was actually cheaper at that time, and it’s still competitive for these kinds of quality metrics,” he added.
According to the contract solicitation, issued by NCI contractor SAIC-Frederick, the successful bidder will provide targeted bidirectional sequencing using “very small amounts of genomic DNA as the starting material,” and will deposit the data to an NCI site, as well as submit sequence trace files to the NCBI’s trace database.
The minimum length of submitted sequences is 300 high quality base pairs, “as determined by an algorithm based on Phred scores (Phred 30),” according to the document. Bidders who did not use Phred scores had to include the algorithm of the accuracy rate in their proposal. The solicitation also states that “any sequencing methods meeting the quality criteria and length of sequence generated may be proposed.”
“I think the next-gen [sequencers] could produce that quality, but it would require a [high degree] of coverage and overlapping sequences,” which would not be cost-competitive with Agencourt’s Sanger pipeline at present, Suh said. “As of right now, I don’t think [the technology] will change. But we are always looking into driving the cost down further,” he said.
Financial terms of the contract were not disclosed.
For the project, the NCI will create a central repository to provide Agencourt with up to 186 independent samples of human genomic DNA at a time, submitted in 96-well plates. The company will also receive “on a periodic basis” — for example, every quarter — a list of genes or genomic regions to be sequenced, according to the solicitation.
NCI has not yet determined how many total samples Agencourt will get, or how many genes it will be asked to sequence, Suh said.
As part of the contract, Agencourt is also developing a website through which project collaborators can access the data.
“The cost of sequencing with Sanger was actually cheaper at that time, and it’s still competitive for these kinds of quality metrics.”
According to Daniela Gerhard, director of NCI’s Office of Cancer Genomics, researchers involved in a number of specific OCG-supported projects have access to Agencourt’s sequencing service.
One such program is the Childhood Cancer Therapeutically Applicable Research to Generate Effective Treatments initiative. According to the NCI’s website, the project, also known as TARGET, is a public-private partnership designed “to identify and validate therapeutic targets so that new, more effective treatments can be developed for children with cancer.”
The initiative resulted from a 2005 workshop sponsored by NCI and the American Cancer Society and “builds upon the experience and expertise NCI has gained” in developing the Cancer Genome Atlas pilot project in collaboration with the National Human Genome Research Institute.
TARGET will focus on gene resequencing “to identify genes that are consistently altered in specific childhood cancers, as their genes represent strong candidates for therapeutic targeting.”
It will also use high-throughput array-based technologies to generate genomic and transcriptomic profiles for selected childhood cancers, and employ high-throughput RNA interference and small-molecule screening to identify and validate therapeutic targets.
According to the website, a pilot project involving researchers at the University of New Mexico, St. Jude Children’s Research Hospital, and NCI will focus on identifying targets for high-risk acute lymphoblastic leukemia. Researchers are planning to select approximately 200 genes for resequencing for the pilot, based on high-resolution genomic and transcriptomic profiles.
This is not Agencourt’s first contract for NCI, nor its first shot at cancer genomics. Under a 2002 contract that was extended two years later, the company provided services for the Mammalian Gene Collection, a trans-NIH NCI-led project to create full-length open reading frame clones for human, mouse, and rat genes.
Agencourt also sequenced approximately 13,000 genes in 22 cancer samples for a large-scale study performed by Johns Hopkins University researchers who published their findings last year (see In Sequence 11/6/2007).