Skip to main content
Premium Trial:

Request an Annual Quote

Euformatics Seeks to Build NGS Informatics Business Around Variant Interpretation, QC Tools


NEW YORK (GenomeWeb) – Bioinformatics firm Euformatics is seeking to build a business in the next-generation sequencing space around its proprietary variant interpretation and data quality management software.

The Finnish firm sells its solutions to hospitals, academic institutions, and molecular diagnostics laboratories. Its OmnomicsNGS solution lets users annotate and interpret variants found in patient samples as well as generate clinical reports. Its interpretations are based on information from many of the same public resources, including databases maintained by the National Center for Biotechnology Information and the European Bioinformatics Institute, used by other companies.

The company's second product, OmnomicsQ, helps users manage the quality of their analysis workflows and ensure that their activities meet best practices and guidelines from the College of American Pathologists, EuroGentest, and other regulatory bodies. Last December, Euformatics said that it had formed a partnership with two third-party quality assessment scheme providers — the European Molecular Quality Network and the United Kingdom National External Quality Assessment Service for Molecular Genetics — which had signed up to use OmnomicsQ in their efforts to develop quality assessment schemes for NGS technology

This week, the company said that it had received CE marking for OmnomicsNGS. While CE marking is not required for NGS software at this time, Euformatics pursued the certification to provide clinical customers with a measure of confidence in the software, Christophe Roos, the company's chief scientific officer and co-founder, explained to GenomeWeb. 

"When we go to a hospital, we seldom meet any IT [experts] and [these users] feel more comfortable if they know that the product has gone through some kind of [third-party] approval or that it is regulated somehow because they are not capable of evaluating it themselves," he said. "It gives confidence that the tool will perform certain things in a certain way and in a consistent way." It's also something of a preemptive move because it's likely that in the near future clinical NGS software companies will be required to obtain the mark for their software, he added.

Euformatics opened its doors about five years ago with an initial focus on the microarray-analysis market before moving into the next-generation sequencing space as the technology's popularity and use surged in research and now clinical contexts. It still has some microarray customers that it supports, but most clients have made the switch to NGS, Roos said.

Both solutions leverage technology and expertise from a now defunct company called Medicel, a firm at which some current Euformatics employees previously worked. "Our main focus has been high-throughput data in research and then in an applied context like in diagnostics and other clinical research," Roos said. "In the beginning it was microarray analysis, but then came the NGS explosion and ... we have gone from high-throughput data to high-throughput sequencing data."

The company chose to focus on variant interpretation because it believed that earlier portions of the sequence analysis process could be adequately addressed by existing open-source software developed in academic environments around the world. "That's an area where we would be competing with bioinformaticians," Roos said. "There is huge amount of intellectual capacity in that area, and we [realized] that it's not possible for us as a company to compete in the academic world."

However, there's still plenty to do post alignment and post variant calling, and more than enough room for an informatics company to stake a claim. Exome-sequencing tests and gene panels yield several thousand variants that clinicians need to be able to analyze and make sense of. "It's something you need help with. You don't do that manually with Excel," Roos said.

"The approaches are different if you are talking to someone doing cancer diagnostics or someone doing rare disease diagnostics or Mendelian inheritance ... there is a need for multiple different approaches," he added.

The company launched its variant interpretation software about two years ago and then added OmnomicsQ last year in response to a perceived need in the market for quality-control tools as NGS gains traction in the clinical space. "There are authorities coming into the picture and saying, 'You are allowed to do genetic testing, but you need to be accredited, and in order to be accredited, you need to fulfil [and] abide by certain rules,'" Roos said. "[The] accreditation was one thing where we realized that people didn't really have tools or understanding to do things by themselves, and so they needed help."

Furthermore, applications of NGS are broadening and creating a new set of quality-control needs that OmnomicsQ can help tackle, Roos added. For example, diagnostics labs developing new tests need to be ensure that they are covering the genes that are considered pertinent to their diseases of interest, and they also need to update their workflows and pipelines as new information is added to the biomedical knowledgebase.

OmnomicsQ is different from traditional laboratory information management systems, which Roos said are designed primarily for storing data but not for calculating quality metrics and running analytics on NGS data and workflows. "NGS technology is based on statistical procedures ... and you need to have very good tools in order to evaluate these [procedures]," he said. "Some sophisticated LIMS will provide some basic analysis tools but not for NGS. [It] is just a storage place."

For now, the company's software has a general focus, but Euformatics has begun exploring more specialized applications for particular disease areas that will be based on its software, Roos said. For both of its products, Euformatics offers per-sample pricing and licensing options. On a per-sample basis, for its quality-control solution the company charges $5 per sample. Per-sample pricing for its variant interpretation product varies because it depends on what sort of testing was performed on the sample — whole-exome sequencing versus a gene panel. For bulk users, the company offers licenses that start at €500 euros (about $551) per month to perform quality control on an unlimited number of samples, Roos said. As with per-sample costs for OmnomicsNGS, license prices for the company's variant-interpretation solution are negotiable for bulk users.

Generally speaking, customers have the option to access the company's software remotely or run them on local servers. The company can also adapt the solution to work within customers' IT environments and existing software, such as LIMS, so that clinicians can pull in patient information from other systems, Roos said.

Specifically, the quality-control software, which can be sold as a standalone or made available as a module in the OmnomicsNGS, is installed and runs locally on customers' laptops but sends quality-control metrics to the Euformatics cloud. "The architecture is [set up] like this because most of the places will not be fine about sending patient data somewhere else," Roos explained. Where the variant interpretation software is installed depends on the institution. For institutions that prefer not to have any patient data on public clouds like Amazon because of security concerns, the company will install the solution behind their firewalls, he said.

Euformatics competes with companies such as MediSapiens, also based in Finland, which also offers analysis software for clinical use. The company initially launched with a focus on oncology but then said in 2014 that it was starting to look into other personalized medicine applications. Recently, MediSapiens was named a partner in an international consortium headed by the Competence Centre on Health Technologies of Estonia (CCHT), which aims to develop personalized medicine applications that combine patients' genomic and health data.