NEW YORK (GenomeWeb) – Swiss bioinformatics company Saphetor said this week that it will provide next-generation sequencing analysis services to Geneva University Hospitals' (HUG) genetic medicine division.
Launched in 2014 and based in Lausanne, Saphetor offers a software-as-a-service platform that automatically annotates and classifies genetic variants from NGS data and is designed to help clinicians quickly and accurately diagnose disease and select the most effective therapies, in particular for cancer and rare genetic disorders. "We focus on processing large datasets, usually exomes and whole genomes, which are only now entering clinical use," Saphetor CEO Andreas Massouras told GenomeWeb. "These large datasets benefit from fast and accurate annotation with our integrated data more than the smaller ones — 1 to 5 genes — that are typically used in the clinic."
The company's platform combines open-source and proprietary algorithms to process raw sequence and call variants in customer samples as well as to interpret and annotate lists of variants. It features a knowledgebase that aggregates information from more than 20 public and proprietary databases, including the Human Gene Mutation Database, ClinVar, and 1000 Genomes. Customers can search for variant annotations directly through the platform's interface or access that information using the company's application programming interface. The system classifies variants based on the five-class score suggested by the Association for Clinical Genetic Science guidelines and links them to information on known disease associations, population statistics, treatment options, clinical trials, and more.
Customer reports provide tables of variants along with their classification and relevant annotations, information on common variants shared by samples submitted as part of the same cohort, details on biological context, as well as data quality and coverage reports.
Under the terms of its agreement with the company, HUG will deploy Saphetor's software and data solution on internal infrastructure and use it in clinical applications focused on diagnosing developmental, neurological, and other disorders, Massouras told GenomeWeb. Saphetor also has agreements with research departments at the University of Geneva to use the solution to study developmental disorders and cancer, he said.
Earlier this year, the 10-person company inked a deal with Lausanne University Hospital's (CHUV) department of endocrinology, diabetes, and metabolism. Researchers there are using the platform in research and clinical applications aimed at diagnosing and treating endocrine disorders — at least one CHUV project that uses the Saphetor platform explores the genetic basis of gonadotropin-releasing hormone deficiencies.
In addition to its commercial product, Saphetor also offers a free version of its knowledgebase called Varsome, which provides access to a subset of the genomic variants that are contained in the commercial iteration of the platform. Varsome, now available as a beta version, covers only the variants that the company has gleaned from public databases such as dbSNP, RefSeq, the Exome Aggregation Consortium, and the Drug-Gene Interaction Database maintained by researchers by Washington University at St. Louis. Making the database freely available offers potential customers an opportunity to explore some of the value that the company can bring to their projects, Massouras noted. Users can help curate variants in the database by adding written comments to the records, and the company plans to add some new features moving forward that will enable more direct variant curation, he said.
A software engineer by training, Massouras transitioned from the finance industry to genomics because of the opportunities in the space enabled by the development of newer and more effective technologies and tools. "It was a unique opportunity in time to do something about it and be part of this revolution," he told GenomeWeb. On the clinical side, in particular, the advantages of using technologies such as whole-exome sequencing to diagnose disease and improve patient care and outcomes were especially clear to him, he said.
At the same time, "we have to [integrate a] vast amount of data from [multiple] sources in order to effectively help the physician interpret the results as more and more variants are produced for patients," Massouras added. He and his team grew frustrated with having to visit multiple websites to locate relevant information on variants and with spending time adapting queries to the particular format and notation used by each site, he said. That led them to focus on creating a one-stop shop for comprehensive access to variant annotations. According to numbers provided by the company, the knowledgebase that underlies its commercial platform contains more than 18 billion variant and gene annotations while Varsome contains about 14 billion annotations.
In Varsome, "you can find information that is already discoverable in public websites, but we have aggregated it in one place and made it easy to query with different formats that geneticists like to use," Massouras said. "So you can query the same variant in different ways ... and it gives you data that you would have to go to 10 websites to find by yourself."
Saphetor's solution competes with offerings from companies such as Sophia Genetics. The Sophia Data Driven Medicine platform provides tools for processing, analyzing, and reporting the results of next-generation sequencing-based clinical tests. The most recent iteration of Sophia DDM includes new so-called artificial intelligence technology that the company claims provides more accurate and precise predictions of pathogenic variants in clinical diagnostics. Sophia has signed multiple partnerships in recent months that couple its software with diagnostic kits from developers such as Devyser and Multiplicom.
Saphetor also competes with Tute Genomics, whose platform includes proprietary variant annotation technology as well as genome interpretation software and an accompanying knowledgebase that was initially developed by Knome; and with Omicia, whose Opal Clinical platform offers proprietary algorithms and a database of gene-variant information for interpreting sequence data in clinical contexts.
But Saphetor believes that it has a unique ability to couple variant data integration with annotation that sets it apart from competing firms. Its approach draws on Massouras' previous experiences in the financial industry, where he was responsible for integrating large quantities of financial data. "Our users and customers have commented that we include more relevant information than our competitors and we are particularly good at cross-referencing it," he said.
Saphetor's plans for the near future include launching the first official version of Varsome. The company is currently developing and adding a number of new features to the platform and is adapting it to the different queries that users want to run. It hopes to release version 1.0 of Varsome in the next month. The firm is also currently in talks with a number of unnamed partners about the possibility of bundling its product with theirs for the market, Massouras said, and will disclose details about these partnerships as soon as they are finalized. Also, it has joined the Global Alliance for Genomics and Health and is currently looking to participate in the group's Beacon initiative.
The company does not disclose details about pricing for its platform. Massouras said that the cost for access to the SaaS version of the platform varies depending on customer requirements, for example, whether they submit raw sequence for processing or whether they start with a list of variants. It also depends on the sample volume — the higher the volume, the lower the cost per sample. There is also a cost for using the company's API, which is also not disclosed. Customers like HUG, who want the data to remain behind their institutional firewall, can choose to have the solution implemented on their local infrastructure.
Earlier this year, Saphetor raised CHF1.2 million ($1.23 million) in seed financing from unnamed investors, along with an undisclosed amount of non-dilutive funding. The company is using the funds for product and business development, according to Massouras.