NEW YORK (GenomeWeb) – Armed with a $1.8 million Phase II Small Business Innovation Research (SBIR) grant from the National Institutes of Health, Bozeman, Montana-based bioinformatics company Golden Helix is developing and refining proprietary algorithms for detecting copy number variants and other structural variants from next-generation sequencing data in clinical contexts.
According to Andreas Scherer, Golden Helix's president and CEO, the funds will help support the next two years of product development. "With the help of the approved funds, we will be able to take calculated risks pushing the currently existing boundaries of the clinical interpretation of next-gen sequencing to the next level," he said in statement.
This phase II grant follows a roughly $150,000 seed grant from the NIH, which the company used to develop its initial technology for detecting CNVs in NGS data and to bring the product to market, Scherer said in an interview. That product, called VS-CNV, is designed to call CNVs in target regions using data from NGS-based gene panels as well as exome and whole-genome sequence from germline and cancer tests. Users can call CNV events that range from 200 base pair single exons all the way up to chromosomal aneuploidies.
In addition to improving the algorithm's performance, Golden Helix plans to integrate VS-CNV and VSClinical software, which is designed for interpreting clinical variants based on guidelines from the American College of Medical Genetics and Genomics and the Association for Molecular Pathology. Golden Helix launched the software in May this year. This way, clinical researchers will have a single solution for detecting point mutations in addition to more complex variations like CNVs, which are harder computationally to detect directly from NGS data. It offers an alternative to using methods like microarrays and multiplex ligation-dependent probe amplification to elucidate CNVs helping to simplify the clinical testing workflow, according to the company.
"If you just focus on single nucleotide variations … you miss [out] on a massive amount of human variation that exists," Scherer said.
In a recent study published in Genetics in Medicine, researchers from the University of California at San Francisco and Invitae found nearly 3,000 intragenic CNVs affecting 384 genes over-represented in individuals with clinically significant results, particularly those affected with neurological conditions. A separate study from 2017 identified rare copy number changes in two genes that seem to be associated with increased risk for Tourette syndrome. A third study from 2016 found the CNVs were over-represented in individuals with schizophrenia.
"We want to eliminate [that] blind spot [and] we also want to make it affordable because sometimes using methods like MLPA or microarray is price prohibitive," Scherer said. "Our solution is key to that because it allow you to do everything [with] one test."
In addition to improving its CNV detection capabilities, Golden Helix will also work on algorithms for other kinds of complex variations such as translocations and inversions, Scherer said. Golden Helix officially entered the clinical testing market with the launch of its VarSeq software in 2014. Designed with clinical laboratories in mind, VarSeq streamlines the process of annotating, classifying, and filtering variants from NGS pipelines.
In 2015, Golden Helix signed an agreement with PreventionGenetics that allows the latter to use VarSeq as part of its exome sequencing pipeline. And a year later, Golden Helix launched VSWarehouse, a database infrastructure for VarSeq that is designed to help clinical research labs better manage and query NGS variant call sets, clinical reports, and variant assessment catalogs. It includes versioning capabilities that let users track changes, and it offers access to manually curated information from repositories such as ClinVar.
The company continues to support the research market's analysis needs through its SNP & Variation suite, which is designed for analyzing and visualizing genomic and phenotypic data. However, there is a growing market for Golden Helix's clinical products, according to Scherer, and he expects that this market will continue to grow once the company adds functionality for detecting complex variations. There are differences in the algorithm for use with germline versus somatic data, but other than that, "our approach is completely agnostic to the area [in which] they are being used," he said.
Golden Helix also plans to roll out improvements to its algorithms over the next 12 months and intends to collect feedback from customers and make improvements to the software based on their comments.