NEW YORK (GenomeWeb) – Seven Bridges Genomics UK will use a £1.95 million (about $2.9 million) grant from Genomics England's Small Business Research Initiative to complete and commercialize a population-scale graph-based genome reference tool and a suite of related software for assembling and analyzing whole genome samples.
The London subsidiary of Seven Bridges — which also has offices in the US and Serbia — is one of five companies selected from an initial pool of 10 to receive grants from Genomics England under the second phase of its SBRI initiative to assess technologies for genomic sequence data analysis and interpretation.
The other companies chosen to move on to the second phase are Cogenica, Genomics Ltd, Omixon, and Oxford Gene Technology. The SBRI awards are managed by Genomics England to support the implementation of the 100,000 Genomes Project, an effort to sequence 100,000 whole genomes from NHS patients by 2017.
Seven Bridges CEO Deniz Kural told GenomeWeb that the company has largely completed scientific and technical development on the core platform technology and that it is preparing to launch two commercial services based on its graph reference tool and analysis suite. The first of these will be a so-called whole genome determination service that will reconstruct graph genomes from raw sequence based on the company's reference. The service will accept data from Fastq, BCL, or BAM files as input and return fully assembled genomes along with information on genetic variants and their respective locations in the genome.
Seven Bridges plans to launch an early version of this particular service in the first half of this year. To that end, the company will begin accepting applications for early access customers at the end of April, Kural said.
A second offering that will follow shortly, based on the same infrastructure, will be a cancer genome service that will provide a primary tumor graph of a cancer sample that will trace the evolution of the tumor over time and provide information about the primary tumor as well as any subclones, Kural said.
Pricing for both services has not been determined.
The Genomics England funds will support some development work including adding security and privacy features to the cloud-based infrastructure as well as scaling up its pipelines to better support the large quantities of data that will flow from the UK project as well as the infrastructure that Genomics England puts in place for the effort, Kural said.
When it launches, Seven Bridges' Graph Genome product will provide researchers with tools to store and analyze large sets of human genomic information faster and more cheaply, according to the company. Customers will also have access to the company's tool suite which will include a graph aligner, a branch genotyper and local assembler, a graph database, a genome browser, and a number of disease-focused applications.
One of the chief benefits of the solution is its improved accuracy. The system boasts a self-improving algorithm that learns as data is analyzed so that its performance improves over time and each new analysis is more accurate than the last. Internally, Seven Bridges has run its technology on hundreds of genomes, Kural said. The company intends to publish scientific papers this year that will provide the fruits of those studies and demonstrate how its solutions improve on the current state of the art.
Improved accuracy translates to reduced costs. Kural noted that access to a more accurate reference means that researchers can actually sequence samples at lower coverage and still be able to call variants as accurately as they would be able to with higher coverage. Because the system learns and improves with repeated analyses of various samples, a user can be confident that if it indicates, for example, that there is a 200 base pair deletion at a particular position in the genome, that the deletion in fact does exist. "You can get away with less sequencing because overall you are more accurate," Kural said.
Moreover, researchers will be able to cut down on storage and computation costs because of the compact way the Seven Bridges offering represents genomes, which makes them much easier to store and compare than is currently possible. "In the past, people needed to store raw BAM files to compare the genomes," Kural said. "We are compressing all that information into a graph so that’s its significantly less expensive to store the genomes."
Being selected for phase two is a real vote of confidence for Seven Bridges and further validation of its efforts in the bioinformatics space, according to Kural. In addition to its work with Genomics England, Seven Bridges also secured one of the National Cancer Institute's cloud pilot contracts, one of three groups to do so. It's also both the smallest and only commercial vendor selected for the initiative.
"We were able to develop the Graph Genome with support from Genomics England," he said. He also credited Seven Bridges’ UK subsidiary for its efforts to push the project the forward. "We are really excited about using what will be the largest population genomics study... to construct a population reference graph for the UK. The efficiency and accuracy improvements of the graph what make national-scale projects possible."
The company also participates in the efforts of the data working group of the Global Alliance for Genomics and Health to develop standards for better representing and including different kinds of genetic variation in the canonical human reference genome, which is maintained by the Genome Reference Consortium.