Skip to main content
Premium Trial:

Request an Annual Quote

Wash U Develops Custom Informatics Workstation to Support Clinical Cancer Sequencing

Premium

This article has been updated to correct the previously reported name of the tool. Its the Clinical Genomicist Workstation not Workbench.

Researchers in the Genomics and Pathology Services laboratory of Washington University, St. Louis, have developed software called the Clinical Genomicist Workstation that provides applications for processing, storing, and reporting the results of clinical next-generation sequencing-based tests.

The workstation was the subject of a presentation given by [email protected] researchers at the Summit on Translational Bioinformatics held last week in San Francisco, Calif.

Developers of the system are exploring options for providing it more broadly under a commercial model in summer 2013, but have not yet settled on pricing details.

CGW offers tools for generating genetic reports that contain pathologist-curated clinical interpretations of genetic variants identified in tested patients' samples. The workstation helps clinical researchers analyze genes and variants in the context of specific clinical indications and public sources of information such as the Single Nucleotide Polymorphism database, the 1000 Genomes Project, and the Catalogue of Somatic Mutations. It can also be integrated with computerized order entry interfaces that physicians use to order tests, as well as laboratory information management systems.

In an interview prior to the TBI summit, Ian Hagemann, a molecular genetic pathology fellow at WUSTL and one of CGW's developers, told BioInform that his lab built the workstation to support the Washington University Cancer Mutation panel, an NGS-based comprehensive cancer panel that tests for 40 cancer-related genes including KRAS, EGFR, and TP53.

He said that WUSTL's GPS lab needed an informatics solution that would enable it to use NGS-based testing in a clinical setting.

The GPS cancer panel, Hagemann explained, is performed on the Illumina platform. It covers all exons "and a little bit of flanking intronic coverage" of the 40 cancer-related genes. Its goal is to identify "variants of either prognostic or predictive significance that have been documented in the medical literature" in order to help clinicians make better treatment decisions, he said.

In addition to identifying clinically actionable variants, "we [wanted] to classify the variants into significance levels or grades based on accumulated clinical evidence" in order to indicate whether patients with particular mutations would have a better or worse prognosis, he explained.

The group has also built an annotated clinical-grade database that includes mutations that are identified frequently in oncology samples, so that rather than "creating a new description or clinical interpretation each time we find that variant, we are maintaining and curating a database of interpretations with a consistent standard as to what constitutes clinical-grade evidence," he said.

Mukesh Sharma, the biomedical informatics project manager in WUSTL's Center for Biomedical Informatics, told BioInform that [email protected] developed CGW in 2011 in collaboration with SemanticBits, a software development company for the health and life sciences industries with offices in Virginia and India.

Sharma, who is also one of the tool's creators, said that prior to building the CGW, [email protected] considered solutions from several vendors but was unable to find a solution that allowed it to analyze, interpret, and present NGS data in a format that pathologists and clinical genomicists could understand and also to generate reports that clinicians could use to determine the best treatments for their patients.

It was also important to have software that could support a clinical rather than a research workflow, Hagemann said. That meant developing a tool that was "role oriented [and] highly customizable" because "we wanted to have the ability to shape the information system to match the workflow of our clinical test," while also maintaining control over the reporting, he said.

Currently, the WUSTL lab is only using the CGW for data from its cancer panel, but it plans to use it to support more tests including clinical whole exome tests for cancer and other constitutional disorders, Hagemann said.

"We also support a number of [undisclosed] clinical trials and translational research projects by next-generation sequencing and we are using the CGW to support those, as well," he said. These projects are studying genes and mutations involved in neurodegenerative diseases and other conditions, he said.

[email protected] plans to sell licenses to CGW for a yet-to-be-determined cost, Rakesh Nagarajan, [email protected]'s director of biomedical informatics, told BioInform.

"The basic premise is a HIPAA-compliant hosting solution and possibly an appliance as options where there is a base annual licensing cost and per-case storage and compute costs," he said in an email.

"The latter will vary based on the level of annotations/analytics that are purchased on a per-panel basis," he explained. For example, he said, customers could purchase a targeted deep sequencing cancer panel along with analytics, PubMed annotations, and curated clinical interpretations; or they could order a cancer tumor/normal exome with analytics and annotations from databases such as COSMIC. "Depending on what analytics and what level of annotations you want, the cost per case will be different."

Sharma explained that CGW's infrastructure allows users to add their own analytics or they can use tools that are already in the workstation for common use cases, such as deep targeted sequencing, exome sequencing, or constitutional exome sequencing.

Users can also "load annotation databases [such as] dbSNP and COSMIC, and create rules to classify variants using these databases for any panel/set of genes," he said.

A potential competitor for the CGW is the Geneticist Assistant Workbench, a software tool with similar capabilities that was launched last week by SoftGenetics. The company developed the tool in collaboration with Mayo Clinic to provide clinical laboratories with a way to track, review, store, and interpret the results of NGS-based tests (BI 3/22/2013). The company is not disclosing how much it charges for licenses.

Another competing system is the GeneInsight suite, which was developed by a team at Harvard Medical School-Partners HealthCare Center for Personalized Genetic Medicine that provides tools for genetic testing labs to store and manage genetic variant information; create interpretative reports; and transmits results to clinicians (BI 8/26/2011).