Appistry is building a commercial version of an informatics pipeline developed and used by researchers in the National Institutes of Health's Undiagnosed Diseases Program to identify the genetic mutations that underlie rare diseases.
Researchers in the trans-division clinical research program — which the NIH established in 2008 — developed the so-called diploid aligner to assemble and compare genomic information in its studies, which focus on individuals suffering from rarely seen health conditions such as Joubert syndrome, progressive supranuclear palsy, and telomerase deficiency and their families. Rare diseases, as defined in the Orphan Drug Act, affect fewer than 200,000 individuals in the US. Of the individuals that are known to have conditions categorized as rare, between 150 to 170 qualify each year for the UDP. The program often sees patients with diseases occurring in fewer than 50 people worldwide.
According to William Gahl, NHGRI clinical director and UDP's director, the diploid aligner's approach offers an alternative to analysis methods that locate harmful variants by comparing an individual's genome to a "generic reference." The problem with this sort of direct comparison, he explained in a statement, is that if both genomes are significantly different from each other, the result of the comparison is an "unnecessarily large number of possible genetic changes." Pursuing each variant to determine which ones matter in the context of the condition in question is both "time-consuming and computationally intensive."
UDP's method does ultimately compare individual data to a standard reference but it's the second half of a two-part process. Initially, it uses genomic information from the parents in the family trio to build a smaller reference dataset that serves as the template for constructing the genome of the child. After the genomes have been assembled, the data from the trio is then compared to a standard reference genome. This approach makes it possible to spot locations where the child's genome differs from its parents and also to make sure that these atypical variants are truly unique to the child and not the result of missed calls or false positives/negatives.
Under the terms of its agreement with the UDP, Appistry is building what the company's CEO Kevin Haar described as a "production quality, scalable" tool for assembling and comparing genomes within a family based on the aforementioned approach that makes use of Ayrris/Bio, Appistry's cloud-based platform for next-generation sequence data analysis.
Cornelius Boerkoel, director of the UDP Translational Laboratory at the National Human Genome Research Institute, told BioInform via email that the UDP selected Appistry as its commercialization partner in part because the company has "worked extensively with the software that we use for process management and could readily integrate their workflow into ours," and because "they exhibited much dedication and commitment to development of new processes."
This partnership is similar in some respects to the one Appistry signed last October with the Broad Institute that gave it the right to commercialize the institute's widely-used Genome Analysis Toolkit (GATK). The Broad picked the St Louis-based cloud computing firm to distribute licenses and support a commercial version of the GATK to provide an alternative solution for those of its users who wanted more in-depth support than the Broad's limited resources could manage, and for those working in restricted environments, such as CLIA-certified laboratories, that cannot change their implementations of the GATK when the Broad updates the system.
Appistry expects to launch the diploid aligner commercially later this year. Michael Groner, Appistry's VP of engineering and chief architect, told BioInform that the partners are making final changes to the pipeline — including updating old versions of tools with new ones — before testing and validating it. Appistry hasn't disclosed how much customers will have to pay to use the pipeline and it is still mulling whether or not to rename the diploid aligner, according to Haar.
Meanwhile, the company has signed a one-year contract with the UDP to analyze data from some of its existing studies using what will be the commercial version of the diploid aligner. "We have this backlog of exomes that have been analyzed by standard procedures but not using the diploid aligner," Gahl told BioInform "The contract is for them to do that analysis." The financial details of that arrangement are not being disclosed.
However, once the pipeline launches officially, "there is definitely a goal here to push this pipeline beyond the NIH" starting with institutions that partner with the UDP on rare disease studies, and to “build a community across a number of different institutions,” Haar said. There's also potential to apply aspects of the approach to research projects in areas outside rare diseases, he added.
Moving forward, Appistry plans to pursue more partnerships similar to the ones it has with the Broad and the UDP. These sorts of arrangements have proved profitable for the company, and "we'll continue to look at ways in which we can work with teams to build pipelines that have broader usability and … address more specific problems that researchers are facing," Haar told BioInform.