Golden Helix has incorporated two new algorithms into its next-generation sequence analysis software that will help users detect and analyze rare genomic variants in sequence data that may be missed by standard variant detection methods.
The Bozeman, MT-based company partnered with researchers at Baylor College of Medicine to integrate the Combined Multivariate and Collapsing and the Kernel Based Adaptive Cluster methods into the sequence analysis module of its flagship product, SNP & Variation Suite.
Golden Helix said the algorithms, which were developed by Suzanne Leal, a professor in Baylor's department of molecular and human genetics, extend its current variant analysis capabilities because researchers can now assess the combined effect of multiple independent rare and common sequence variants on disease phenotypes.
The sequence analysis module, launched earlier this year, provides tools to import, manage, and manipulate hundreds to millions of common and rare variants to assess their impact on diseases and other traits; find genes or regions with an abundance of variants in samples; assess the rare variant burden; and understand the contributions of rare variants using functional predictions (BI 01/28/2011).
Rare variants are thought to play a larger role in some diseases than common ones, Christophe Lambert, president and CEO of Golden Helix, noted in a statement.
However, standard methods used to analyze common variants and analyze GWAS data are underpowered to test for rare variant complex trait associations, he said.
That’s because these methods typically focus on just one genomic location at a time where variation may be common to only a tiny fraction of the study population, Lambert explained to BioInform. "Whereas if you could aggregate all the positions where there could be deleterious mutations in a gene, then you would essentially be modeling the contribution to disease of that gene as a whole."
Next-generation sequencing is proving adept at discovering more genetic differences than previously possible with genome-wide association studies, and as a result "a different type of analysis is required" like those "embodied" in Leal's methods, he added.
"To be able to look at rare variants and be able to deal with ... confounders basically gives researchers more power to find associations," Lambert said. "That means ... moving forward ultimately explaining the basis for disease and then doing something about it."
Combined Multivariate and Collapsing, or CMC, and Kernel Based Adaptive Cluster, or KBAC, published in 2008 and 2010 respectively, are statistical algorithms that detect associations between complex disease phenotypes and rare DNA sequence variants.
CMC tests for the cumulative effect of variants within specified bins, which are based on allele frequency or other properties that occur in defined genomic regions, usually genes.
KBAC differs from CMC in that both variant classification and association testing are unified into a single procedure. KBAC models the risk associated with multi-site genotypes rather than collapsing individual genotypes based on specified bins.
In addition to providing a common user interface for both algorithms, Golden Helix said it used a regression framework to provide the algorithms with the ability to correct for confounding variables — gender and body mass index, for example — that might skew the results.
Since Leal and her team published the methods, researchers have adapted them for studies on ailments such as breast cancer, preeclampsia, and rheumatoid arthritis.
These methods could also be applied to results from prior GWAS studies to see if additional variants could be found, Lambert said.
Not in a Decline
On the commercial front, Golden Helix is receiving a lot of interest in its sequence-analysis software from groups in academia, government institutions, pharmaceutical companies, and biotech firms, according to Lambert.
He said the company is optimistic that sales of its tools, which can be used by experienced programmers and biologists with some computing expertise, will rise as the NGS market matures.
However, microarrays may also help drive sales, Lambert said, adding that his firm plans to continue to refine and improve its microarray data-analysis software.
He pointed out that companies like Illumina appear to be seeing "a steady volume" of business in the microarray space based on their earnings reports.
Furthermore, "there is also a lot of work in cytogenetics [and] cancer research where people are using microarrays on the research side and also in clinics," he said. "I don’t think microarrays are in a decline yet ... [in fact] you are probably going to see a lot more custom arrays."
Since it launched its NGS module, Golden Helix has hired two employees to provide technical support, quality assurance, and software development, and plans to add to its sales force "as revenues grow," Lambert said.
He declined to comment on how many employees Golden Helix currently has.
Have topics you'd like to see covered in BioInform? Contact the editor at uthomas [at] genomeweb [.] com.