Main

SNP genotyping is both big science and big business; as such, it's no surprise that substantial progress happens where the two intersect. In this vein, collaboration between researchers at ParAllele Bioscience, Inc. and the Baylor College of Medicine, both participants in the ongoing International HapMap Project, has produced a new system for the large-scale, multiplexed analysis of hand-picked SNP arrays (Hardenbol et al., 2005).

“With our technology... you can choose 10,000 of the most relevant SNPs for your particular study,” explains ParAllele researcher Tom Willis, “and make a multiplexed assay that would allow you to score those particular SNPs.” The centerpiece of this technology is the molecular inversion probe (MIP), a reagent previously developed by members of this group (Hardenbol et al., 2003). A single MIP is designed for each SNP; after hybridization, separate enzymatic reactions are carried out to identify any possible nucleotide variant. In the subsequent microarray analysis, SNPs are identified using a MIP-specific, thermodynamically optimized bar-code sequence: for each of the four reactions, bar codes are PCR-amplified in the presence of a different fluorophore, and by analyzing the array data, one can assess the ratio of nucleotide variants at each site.

This system was used to genotype sets of 6,000 and 12,000 SNPs against genomic samples from the HapMap collection; these specimens were collected from families, allowing the accuracy of the process to be confirmed by checking the mendelian inheritance of each SNP. The MIP strategy proved suitable for recognizing more than 85% of the SNPs the researchers selected, and the data consistently achieved levels of accuracy that surpassed 99.6% and were highly repeatable. With less stringent standards, higher SNP recognition was possible, although accuracy was compromised slightly as a result.

Willis indicates that this system is easily scalable and suggests that chips capable of analyzing 100,000 SNPs should be possible. “There are no large obstacles today, other than just building the resource of MIPs and optimizing the label to get sufficient signal off of the arrays.” ParAllele is continuing to develop this technology and participate in the HapMap Consortium, but other applications are also being considered for this system, and Willis sees great potential in the development of chips specific for polymorphisms that affect protein sequence or alter gene expression. “We've developed one panel... that's got 10,000 amino acid–changing SNPs in one assay,” he says. “[And] one obvious way to expand that is to go from just amino acid–changing SNPs to other categories of functional SNPs, like splice-site SNPs and promoter SNPs... We think this represents an important, complementary genetic strategy to HapMap-type approaches.”