Rochester Startup OmniTier Looks to Speed Up Whole-Genome Sequencing

Rochester Startup OmniTier Looks to Speed Up Whole-Genome Sequencing

OmniTier says its new tech platform has big implications for the spread of personalized medicine.

A Rochester startup co-led by a former IBM and Mayo Clinic computer scientist is getting ready to roll out a new technology platform aimed at speeding up and lowering the cost of whole-genome sequencing, seen as a key step toward the widespread adoption of personalized medicine.

OmniTier Inc., which also has offices in Milpitas, California, will introduce its CompStor product in the fourth quarter. It’s touting the platform as a breakthrough in the practice of assembling sequenced genomes “de novo,” or from scratch and without a reference — a process which yields the best and most accurate genetic information but which currently requires supercomputing capabilities and long wait times due to the massive amount of data generated.

CompStor, the company says, uses a unique, tiered memory configuration to accomplish such de novo genome assembly in a matter of minutes and on standard, low-cost servers, thus potentially “opening the door to hundreds of institutions globally that are eager to leverage next-generation sequencing techniques.”

OmniTier’s California-based CEO is Hemant Thapar, a data storage expert, successful entrepreneur and former IBM computer engineer in San Jose.

Its chief technology officer, meanwhile, is Jon Coker, who has a long resume in Rochester stretching back to his first hire there by IBM in 1984. He leads the company’s local office and is also a former computer engineering supervisor for Mayo Clinic.

Coker said the three-year-old firm now has 15 employees and is “actively hiring” as it prepares to roll out CompStor to potential customers including genetic researchers, academic labs and others who are involved in “expediting the path to personalized medicine.”

“This first implementation of CompStor is as a ‘variant calling’ mechanism (in which users search for genetic variations within single genomes),” he told TCB. “But our bigger strategy is to get into scientific computing in general, and that includes analysis of not just one particular genome data set, but millions of such data sets.”

That longer-term strategy, he added, will make the platform useful not only for providers of “high-throughput whole genome sequencing,” but also to clinical human health researchers who are looking to comb through massive troves of data to discover new genetic variations associated with complex human health conditions such as cardiovascular disease and cancer.

The product rollout is coming after the publication of a study conducted by OmniTier and Mayo Clinic’s Center for Individualized Medicine which successfully demonstrated human DNA “variant calling” using de novo assembly techniques could be accomplished in less than two hours.

The solution, the study found, could also be scaled to a dataset size previously considered too large for de novo techniques but which is needed to reliably identify new and rare genetic variants.

Doug Bodde, OmniTier’s director of business development, said that while the firm has no financial or product licensing ties to Mayo Clinic, it has “certainly benefitted from Mayo’s sponsorship and facilitation of the entrepreneurial community here in Rochester. High-performance computing for the life sciences hasn’t been a typical Rochester business, but the opportunity is here and we’re pursuing it.”

After an earlier seed financing, Coker indicated OmniTier is currently out raising a Series A venture capital financing round.