When a newborn baby is suspected of suffering from a rare disease, it may be a race against the clock to identify an actionable genetic cause for the baby’s condition. Data sharing is key to harness the potential of new DNA sequencing technologies and the growing knowledge base. In collaboration with clinical partners, DNV GL has developed Trusted Variant eXchange (TVX) for sharing of variant interpretations between labs.
A rare disease affects fewer than 1 in 2,000, but as there are between 5,000 and 8,000 different rare diseases, collectively they affect more people than cancer and HIV-AIDS combined. 80% of rare diseases are genetic in origin. Modern DNA sequencing technologies offer a unique opportunity to identify this genetic variation and reduce the time to reach a more precise diagnosis and initiate management.
Finding the genetic variant causative to a disease among the 3 billion letters compromising a person's genome is like finding a needle in a haystack. Bioinformatic tools are used to filter out common variations in the normal population and zoom in on genes relevant to the disease. The most labour-intensive part of the process is to investigate the up to a couple of hundred remaining candidate variants and interpret their relevance to the disease (‘classification’), using all experience and knowledge available. Variant identification and interpretation is not a standardized process, and even as guidelines are starting to be adopted, the outcome will depend on specific competence and experience, local interpretation of criteria, access to data, etc.
Benchmarking demonstrates that variant classifications available from the largest international databases vary in quality. Evidence may be lacking or incomplete, and classifications may be inconsistent. A pathogenicity analysis of the ClinVar database found that a majority consensus was reached for only 89.3% of submitted variant classifications. Evidence accrues continually, and outliers are more frequently found with older classifications. Sharing of variant classifications is critical for quality assurance and harmonisation to improve clinical decision support and patient safety, but this information is currently not systematically shared between clinical genetic departments.
Sharing of variant classifications will only improve the diagnostic process if they are trusted, and if sharing and access is integrated in the clinical work flow. By working with partners in the BigMed project and gathering needs from clinical genetic laboratories in the Nordic countries through the Nordic Alliance for Clinical Genomics, DNV GL has developed the Trusted Variant eXchange that facilitates sharing of quality assured variant classifications with trusted partners of choice.
Key features of TVX include:
- Quality assurance of variant classifications.
- Controlled and secure sharing of variant data with trusted partners of choice.
- Laboratory specific dashboards that include details of classification discordance.
- Push notification of discordance of new and historically classified variants.
Sharing of classifications between laboratories will lead to harmonization of classification procedures and standardization of formats used, driving continuous improvement of this emerging area of diagnostics.
Your genome encodes the instructions for making and maintaining you. It is written in a chemical code called DNA. Your genome is all 3.2 billion letters of your DNA. It contains around 20,000 genes. Genes are the instructions for making the proteins our bodies are built of – from keratin in hair and fingernails to antibody proteins that fight infection.
Sequencing is the technique used to ‘read’ DNA. It finds the order of the letters of DNA (A, T, C and G), one by one. Sequencing a human genome means finding the sequence of someone’s unique 3 billion letters of DNA.
Some rare diseases are caused by as little as a single change (variant), like a spelling mistake, in someone’s DNA. Looking at the genome of a person affected by a rare disease can help find which DNA changes might be causing the problem. For some patients, knowing more about their genome may mean that a particular treatment can be recommended.
Source: Genomics England