Companion diagnostics
A companion diagnostic identifies the patients for whom a particular treatment is appropriate. The analysis therefore has to meet both the clinical requirements of the test and the requirements that apply to a regulated system.
Research and diagnostic work
The methods are familiar. The validation and reporting requirements are different.
Validation
Sensitivity, specificity, precision, and limit of detection are established for the variants or markers covered by the test and for the specimen types used for the test.
Intended use
The test is validated for a defined population, specimen type, and therapeutic question. Results outside that scope are outside the validated use of the test.
Traceability
Each result can be traced to the pipeline version, reference data, parameters, and input files used to produce it.
Changes to the pipeline
Changes to the analysis are assessed before they are made. Depending on the change, that can mean revalidation and updated documentation.
The validated pipeline
A research pipeline may change as the method changes. For a diagnostic pipeline, the tools, parameters, and reference data are tied to the version that was validated.
Changes to those components go through change control before they are made.
The report is used by an oncologist making a treatment decision. It shows the result, what was assessed, what was not assessed, and the limits of the assay.
If those points are unclear, the report can create a clinical safety issue.
Our genomics work includes somatic variant calling and biomarker analysis using BWA, GATK, and SAMtools. We also build software for digital pathology, including biomarker scoring on whole slide images. That work sits alongside our validation and traceability work for regulated software.
Biomarker scoring, an interactive slide viewer, Python and Playwright automation, and end to end QA for an oncology digital pathology platform.
Read the case study →