Perlegen Sciences will be among the first to get a glimpse of an updated version of 454 Life Sciences’ sequencing platform, which will offer longer reads and higher throughput than the current FLX system, In Sequence has learned.
The new technology, which 454 plans to launch next summer as a system upgrade, will be used in a pharmacogenomics collaboration the two companies announced last week (see In Sequence 10/2/2007).
In the study, Perlegen and Roche subsidiary 454 will sequence a number of candidate regions that are associated with side effects in several hundred patients who were treated with thiazolidinediones, a class of insulin-sensitizing drugs. Perlegen previously identified the regions through a genome-wide association study involving several thousand patients who received the drugs.
The aim of the project is to create a clinical test, which might be sequencing-based, that predicts whether patients are likely to experience the side effects, which include fluid retention and its off-shoots.
454 has said for a least a year that it is working on an upgrade to its technology that will provide longer reads. The current Genome Sequencer FLX version produces read lengths of between 200 and 300 base pairs.
The new reads, at least 400 base pairs in length, will be part of a “broader system upgrade” that will increase the platform’s overall throughput, according to Tim Harkins, Roche’s marketing manager for the 454 sequencing platform.
The upgrade, which will affect reagents and software but not the instrument itself, is expected to increase the platform’s throughput to 1 gigabase of data in 24 hours, both through longer reads and more reads per run. At present, the FLX generates up to 100 megabases of data, or 400,000 reads, in an eight-hour run.
The company anticipates that the cost per run will stay the same, which would effectively reduce the GS FLX’s per-base cost.
454 also plans to use its new multiplexing technology in the project, which allows it to sequence at least 96 samples in a pool. Later this fall, the company plans to release a first set of 12 tags to customers, as well as sequence information for 96 tags (see In Sequence 8/8/2007).
For 454, the collaboration serves “to demonstrate the power of a combined technology approach for human resequencing,” according to Harkins. He added that for this application, “we found it necessary to use the system's extended read length of 400 base pairs.”
Although this is its first collaboration with 454, Perlegen has been a 454 customer since the end of last year and owns two GS FLX instruments, Bryan Walser, the company’s interim CEO, told In Sequence.
In June, Perlegen said it had collected more than 3,000 DNA samples from diabetic patients who were treated with Takeda’s Actos and GlaxoSmithKline’s Avandia, two thiazolidinediones used to treat type 2 diabetes.
These drugs, also known as TZDs, can cause fluid retention in up to 15 percent of patients, a condition that results in edema, weight gain, and sometimes congestive heart failure. Perlegen has licensed a “phase III ready” TZD from Mitsubishi Pharma.
According to its website, Perlegen has already conducted a genome-wide association study of about 2,300 patients using its high-density microarray platform, and found a six-SNP haplotype associated with patients who developed edema. Three of those SNPs resulted in amino acid changes in a single protein.
Walser said the sequencing collaboration with 454 is “a continuation of that study” that will allow the researchers to find even infrequent genetic variations that contribute to an increased risk for side effects.
“The whole-genome scanning arrays are very important and interesting, both for the direct information they can convey and for the information they can convey about what genes might be involved,” Walser told In Sequence last week. But only sequencing can help scientists understand the fine variation, he said.
“When you look at a whole-genome association study using a standard array or chip, it’s more or less looking at mile markers along the freeway,” Harkins explained. “What you don’t know is what’s between those mile markers.”
“When you look at a whole-genome association study using a standard array or chip, it’s more or less looking at mile markers along the freeway,” Roche’s Harkins added. “What you don’t know is what’s between those mile markers,” and sequencing can provide this information.
Both Perlegen, using its current GS FLX machines, and 454, on its upgraded instruments, will contribute sequencing data to the project, and 454’s involvement is speeding up the project.
“Having 454 pick up some of it and run it through their high-throughput boxes lets us finish by the end of the year,” Walser said.
Perlegen is also contributing its proprietary sample prep and amplification technologies that allow it to select specific regions of the genome for sequencing. The company is in the midst of patenting the technology, and Walser would not divulge any details about it. Asked whether it could run 50,000 reactions in parallel, he said, “we are better than that.”
Perlegen will have the rights to any genetic markers or clinical test that might come out of the study. Such a test, which would be used to determine which patients are likely to develop side effects from TZDs, would likely be sequencing-based, according to Walser.
“You never know where technology is going to lead you, but if what you are looking at is variation in the gene itself, not a specific variation, but whether the gene is variable versus a reference standard, … then you are always going to be stuck with a sequencing platform.”
Perlegen chose 454’s technology for this project because it found its relatively long reads and “very high accuracy” to be “particularly appropriate for this kind of work,” according to Walser. Stitching together several kilobases of DNA from specific regions of the genome is preferably done with fewer, longer reads than many short reads, he added.
Besides this study, the company is using its two FLX instruments for a variety of projects “where that’s the right technology to make discoveries in genetic variations.”
At the moment, 454’s platform is the only next-generation sequencing technology Perlegen has in house. But the company is keeping an open mind to other technologies. “There are horses for courses,” Walser said. “You want to do a different type of inquiry, you have got to reevaluate the field and understand what’s best for a particular use,” he added. Analyses of structural variation, for example, might warrant a different type of sequencing technology, he added.