Clinicians and researchers face a daunting task when developing gene panels for any particular disease. But rather than research biomarkers and design panels by committee, rational panel design can now leverage disease-gene-variant associations found in scientific evidence from the medical literature.
Many times, gene panels are designed over a number of months by committee and laborious searches for evidence on which genes, exons and hotspots to include in the panel. Given the lack of consistency in gene panels for clinical reference labs on the same disease, there is a lot of subjectivity in the gene and hotspot selection process.
For example, looking across 91 cardiomyopathy panels, a total of 1,968 different genes are targeted. However, of these 1,968 genes, only 1 gene is included in every panel (ACTC1). Furthermore, 20% of these panels do not include the second most widely tested gene (MYH7) on their panel and the top 20 genes are included on these panels only 51% of the time. This doesn’t even address the differences in the testing at the exon and variant level.
At the clinical level, the testing variance is just as diverse. As the figure above shows—the panels four different reference labs only have 27 genes in common with 104 uniquely tested at just one lab.
The clinical diagnostic industry is hurting itself with this inconsistency. Referring doctors and patients are easily confused on which genes need to be tested for a particular disease and often fall back to a lab’s reputation, sales ability, or relationships when selecting the tests. However, sophisticated doctors and patients are left asking “Why?”.
The lack of rational design can also put the clinical labs on the defensive when looking for insurance reimbursement. Payers struggle to understand how one company only needs to screen 44 genes and another is charging to screen 134 genes. Without evidence to support the larger panels, the insurer’s response is to deny payment.
Our customers asked for a more engineering-driven approach to panel design. One that is repeatable and provides the primary evidence required to build a gene panel to screen for a specific disease. At the same time, they didn’t want a software to automatically select the genes and variants for their panel. They want the ability to curate the primary evidence and make the decisions themselves.
So we took on the challenge. As a result, we are now offering Mastermind Panel Design Service. We not only drastically cut the time it takes to create a set of disease panel candidates, we also pprovide literature citations for each and every candidate variant, gene and exon candidate.
We are able to deliver this break-through approach to panel design by mining Mastermind’s comprehensive knowledge base of disease-gene-variant relationships curated from primary medical literature to provide a list of candidate genes and variants that have been associated with a specific disease. Every candidate gene and variant is tagged with a list of every publication that associates the disease to the biomarker. The final panel can be curated from the candidate genes, exons and variants based on the associated literature.
Our scientists provide a report of the data collected from the literature—filtered and sorted to customer specifications. The comprehensive and prioritized lists of genes and variants (coding and non-coding, single nucleotide and small indels, amplifications/deletions and fusion events) are delivered in a tabbed spreadsheet format for final selection and curation by the customer.
Contact us if you’re interested in learning more about our Evidence-Based Panel Design approach.
Updated: Download: White Paper on Evidence-Based NGS Gene Panel Design