By Malorye Branca
It can take millions of genotypes (the scoring of a single nucleotide polymorphism, or SNP, in a single patient) to uncover links between SNPs and disease or drug response genes. One year ago, the average cost was $1 per genotype — beyond the reach of most companies. Now that the price is hovering around 35 cents per genotype for some systems, more candidate gene surveys are within reach. But genome-wide studies, aimed at finding all the SNPs associated with a disease, remain exorbitant.
"We estimated that you would need about 500,000 SNPs to do such studies," says Richard Judson,
Orchid's SNP-IT technology uses a simple three-step process to directly analyze genetic information at the site of a SNP. First the technology identifies the precise area of DNA where the SNP is located (left). Then an enzyme called DNA polymerase catalyzes a reaction that adds a signal specific for the SNP (center). Finally, the result is read out using one of various methods commonly used by researchers (right). This flexibility gives SNP-IT a key commercial advantage. SNP-IT tag array is Orchid's next-generation technology innovation which enables unparalleled flexibility, throughput, and robustness through multiplexing, the ability to perform multiple SNP analyses at one time.
senior vice president of informatics at Genaissance. Other figures are much higher. "Discussions around the NIH's [haplotype] map initiative suggested it could require 1 or 2 million SNPs and about 1,000 patients," he says. This disparity results from the many uncertainties and the different assumptions people make (see July Bio·IT World
, page 46).
A dozen or more companies offer genotyping tools and/or services. Every step — from finding a putative SNP through sample preparation and scoring the genotype — can influence the speed and accuracy of the project. Perhaps the biggest development is the number of validated SNP assays, some of which can be selected and purchased over the Web. The idea is so simple it's a wonder no one thought of it earlier. "We dramatically simplified the steps going from getting a piece of information in a database (i.e., 'this looks like a SNP') to having an assay for the SNP that works in the lab," says Dennis Gilbert, vice president of genomics for Applied Biosystems Group. The company launched its Assays-on-Demand online genotype-product store this summer.
Applied Biosystems has 90,000 validated SNP assays on hand, and is aiming for 200,000 by year's end. Sequenom Inc., which launched its RealSNP.com Web site in April, has approximately 400,000 SNP assays available, more than 200,000 of which are confirmed (common) variants. Orchid BioSciences Inc. and Third Wave Technologies Inc. each have approximately 100,000, and Illumina Inc. has about 10,000.
Other steps are becoming easier too. "Our Autoprimer.com Web site allows you to load a sequence and design primers for PCR and genotyping," says Michael T. Boyce-Jacino, vice president and chief technology officer at Orchid. "For us, validation means it works in your lab," says Gilbert. "And the workflow with TaqMan is simple: You take the DNA and the enzyme, mix [them] together, and it works the same way every time."
Speed and capacity have also changed dramatically thanks to multiplexing (doing multiple reactions at once). "Multiplexing is a term that is used loosely," says Bill Craumer, director of marketing communications at Illumina. "We routinely prepare samples and perform assay amplification at 1,000-target PCR multiplex levels." Each Illumina Sentrix array has up to 384 fiber-optic bundles and can interrogate up to 1,500 targets.
Several companies, including Illumina, Orchid, Third Wave, and Sequenom, are working on very large-scale projects. "Many people are looking to do genomewide scans," says Orchid's Boyce-Jacino. "Our product addresses that." Orchid also recently launched a whole genome scan mapping service. Applied Biosystems, according to Gilbert, is focused primarily on candidate gene analysis. "That customer already has a set or list of genes they want to study SNPs in," he says.
Comparing costs can be difficult. Some systems require purified oligonucleotides, others do not, and Third Wave's process does not require PCR — a key cost consideration. Accuracy also plays a part. "The denominator determines the cost of genotyping," Gilbert says. "If half the genotypes have to be redone, you are not saving money."
As projects move toward the clinic, new considerations will arise. "Pharma companies want to standardize the development of SNP assays," says Boyce-Jacino. "In case it becomes a diagnostic, they don't want to have to switch biochemistry. We have a very robust biochemistry that, thanks to our propagation efforts, is now available on several other platforms as well."
Back to The New, New Pharmacogenomics
|Pushing the Throughput Envelope
|Applied Biosystems Group
||TaqMan — 5' nuclease assay with ABI Prism 7700 or 7900 HT sequence detection system
||200,000 per day on 7900 HT
||BeadArray technology deployed on Sentrix array matrices
||More than 1 million genotypes per day (using 7 matrices)
|Luminex Corp. (with Tm Bioscience Corp.)
||Universal Array platform — xMAP bead system
||240,000 genotypes per day
|Orchid BioSciences Inc.
||SNPstream UHT — SNP-IT tag array technology
||800,000 genotypes per day
||PSQ 96 and PTP systems — sequencing by synthesis
||100,000 genotypes per day (Can be increased through multiplexing)
|Qiagen Genomics Inc.
||Masscode — PCR-based SNP discrimination assay
||65,000 genotypes per day
||MassARRAY 7K, 20K, and 200K systems
||200,000 genotypes per day on the 200K system
|Third Wave Technologies Inc.
||Invader — allele-specific hybridization with novel signal-amplification technology
||500,000 genotypes per day.