Date Published: April 25, 2019
Publisher: Public Library of Science
Author(s): Broder Fredrich, Marcus Schmöhl, Olaf Junge, Sven Gundlach, David Ellinghaus, Arne Pfeufer, Thomas Bettecken, Roman Siddiqui, Andre Franke, Thomas F. Wienker, Marc P. Hoeppner, Michael Krawczak, Ruslan Kalendar.
Massively parallel DNA sequencing of clinical samples holds great promise for the gene-based diagnosis of human inherited diseases because it allows rapid detection of putatively causative mutations at genome-wide level. Without additional evidence complementing their initial bioinformatics evaluation, however, the clinical relevance of such candidate genetic variants often remains unclear. In consequence, dedicated ‘matching’ services have been established in recent years that aim at the discovery of other, comparable case reports to facilitate individual diagnoses. However, legal concerns have been raised about the global sharing of genetic data, particularly in Europe where the recently enacted General Data Protection Regulation EU-2016/679 classifies genetic data as highly sensitive. Hence, unrestricted sharing of genetic data from clinical cases on platforms outside the national jurisdiction increasingly may be perceived as problematic. To allow collaborative data producers, particularly large consortia of diagnostic laboratories, to acknowledge these concerns while still practicing efficient case matching internally, novel tools are required. To this end, we developed VarWatch, an easy-to-deploy and highly scalable case matching software that provides users with comprehensive programmatic tools and a user-friendly interface to fulfil said purpose.
In recent years, high-throughput DNA sequencing of clinical samples has become routine practice in the diagnosis of human inherited diseases with monogenic etiology. In consequence, a rapidly growing number of genetic variants are being discovered for which a causative role in a given disease phenotype may be suspected. The actual establishment of causality is however difficult, particularly for rare diseases, and depends upon either functional experiments or smart and comprehensive data sharing. The former is usually highly demanding and, hence, prohibitive unless the disease in question is clearly defined, sufficiently frequent and has a reliable functional test available. An example of this type of condition is provided by hereditary breast and ovarian cancer (HBOC), where exhaustive artificial mutants of the BRCA1 gene were recently generated by saturated genome editing and successfully tested for their functional effects . Although this method holds great promise, diseases like HBOC are still an exception, rather than the rule, in terms of feasibility. The second approach to infer causality (i.e. the sharing of data) aims at identifying one or more independent cases with an identical or a closely related combination of variant and phenotype. Historically, such exchange of information has been facilitated by word-of-mouth, scientific journals and public databases. Owing to the rapid growth of information generated, however, these traditional routes of communication are becoming more and more inadequate for timely data dissemination and access, particularly with a view to the necessary follow-up of unsolved cases.
With the advent of cost- and time-efficient means of primary data generation, particularly next generation DNA sequencing, turning the available genetic information into predictors of human disease has become a key concern of medical genetics research and care. The actual challenges, however, differ notably between the different types of genetic etiology involved: Whilst common complex phenotypes require transformation of the weak to modest statistical genotype associations observed in large studies into an understanding of biological mechanisms, the small data basis usually characterizing monogenic diseases calls for better means of collaboration and communication. Since many rare genetic diseases follow a simple mode of inheritance, there is reason to hope that an accumulation of independent albeit clinically comparable cases facilitates rapid pinpointing of the underlying genetic mechanisms. Against this background, the emergence of dedicated international matching platforms such as Phenomecentral  as Genematcher  has been an invaluable step forward towards the swift and comprehensive, ideally world-wide, connection of medical genetic case reports.