Helix has developed a “sliding power window” population genomics method that allows researchers to maintain the statistical power of analyses involving numerous rare variants spread throughout a broad population.
In a medRxiv preprint posted last month, the firm used this technique to identify a subset of people with atrial fibrillation and carrying truncated variants of the Titin gene (TTNtv) who were at higher risk for cardiomyopathies and subsequent heart failure than non-carriers.
The company claims that this provides evidence for the first population-level screening method with clinical utility for cardiomyopathies.
Despite a known correlation between TTNtvs and dilated cardiomyopathies, one’s Titin status has typically been excluded from population screening consideration because the truncated variants have variable and often low penetrance.
Nonetheless, the chance of carrying some sort of TTN variant is high relative to other genes, owing to Titin’s size. With 364 exons, Titin is genetically complex and forms the largest human protein, providing ample canvas for variation.
“Titin was [also] a really interesting gene because we saw seven different gene-disease associations with [it] and phenotypes of interest from the electronic health record,” said Kelly Schiabor Barrett, a research scientist at Helix and the study’s first author. “They were all low penetrance but we thought that there might be something more there that warrants further study.”
Using data from the UK Biobank and the Healthy Nevada Project, Barrett and her colleagues discovered that carriers with atrial fibrillation, or Afib, were approximately five times more likely to develop cardiomyopathies than carriers without Afib.
“We found that stratifying the variation according to [patient] phenotypes … helped identify the people who are at the highest risk of having disease,” said Liz Cirulli, a principal scientist at Helix and the study’s co-first author.
In order to identify this carrier subgroup, the Helix researchers used the power window method to improve the statistical power of its analysis by maintaining roughly the same number of people with a rare variant in each window. While sliding window analysis itself is not new, this appears to be the first one developed to slide according to statistical power, rather than by the number of nucleotides or gene variants.
“Say you have one window that has 40 people with a rare variant in that region,” Cirulli said. “You analyze that and then you slide over a little bit until you get a different group of 40 people with rare variants, and you analyze each of those windows separately. And because they all have the same statistical power, you can then compare the metrics that you get from each of those analyses.”
This way, she explained, “you can rank all the windows in the genes to see which ones are the most [disease]-associated and which are the least associated.”
“Conceptually, identifying a region rather than a specific location/variant that is affected by a truncation event makes sense,” Alexander Lachmann, an assistant professor of computational biology at the Mount Sinai Center for Bioinformatics, noted in an email.
“Since all windows have the same 40 [carriers] the number of [cardiomyopathy] and Afib cases are directly comparable,” added Lachmann, who was not involved with the study.
Helix has filed a patent on the method and plans to publish a more detailed paper on it. The firm is also currently using its sliding power window in several other studies, which it hopes to publish over the course of the year.
While no specific product is tied to this research, Helix hopes that this and future studies will drive business by showcasing the company’s capabilities in the population genetics space.
“Having groundbreaking research like this, that really makes the case for more ways that you can do population genetic screening, and better information that you can get back to patients is really a big part of the research strategy that we have,” Cirulli said.
“This study is important because it is the first population-level screening of this topic for better personalized medicine,” Yukihide Momozawa, a team leader at Japan’s Riken Center for Integrative Medical Sciences who was not affiliated with the study, said in an email.
Helix hopes to replicate its findings in additional cohorts and in larger populations, and to see TTNtv screening adopted as a guideline alongside family history and routine monitoring.
Currently, Barrett said, “there are generally no guidelines for screening in the cardiovascular space outside of just a family history.”
If this study’s findings hold, however, Barrett said that “you may be able to use [screening] to find that subgroup of Titin carriers that are at highest risk of disease.”
Helix is currently engaged in several population genomics projects, with collaborations including the Memorial Hermann Health System and the HealthPartners integrated health system.