Artificial Intelligence | Genetic Testing
How we can advance our understanding of gene variants to help improve patient care
Why Invitae submitted over 1 million gene variants to a free, public archive to inform the global scientific community
When you think of data sharing, you may picture companies misusing your information or selling it to other parties. But in genetics, data are the key to unlocking new information that could improve and extend patients’ lives. To truly advance medicine using genetic data, no clinician, lab or organization can work alone.
Working together and sharing anonymous genetic variant information is how we develop breakthroughs.
That’s why of Invitae’s four core founding principles, two focus on data:
- Genetic data are more valuable when shared
- Patients own and control their genetic information
A collective effort to gene variant interpretation
Only through authentic and meaningful data sharing can the field of genetics reach its full potential. Invitae shares all gene changes (variants) and classifications supporting the interpretation of disease-causing variants and variants of uncertain significance.
Among other activities, Invitae has played a leading role in ClinVar and ClinGen, two National Institutes of Health (NIH)-supported programs dedicated to sharing clinical genetic testing data and knowledge to advance patient care.
When sequencing the As, Ts, Gs, and Cs of someone’s genome, the difficulty lies in solving what the millions of variants mean for one person. Many variants are likely not disease-causing (benign). Often, however, we need more evidence to classify a variant confidently.
That’s how we end up with a variant of uncertain significance. But by sharing anonymized variant information on public databases like ClinVar, we help labs and clinicians worldwide to improve our understanding of gene variants.
ClinVar also helps highlight areas for improvement in our collective comprehension of variant interpretation. They offer a free, public archive of clinically evidenced reports that address the relationships between genetic variants using variant submissions throughout the globe. Invitae has been one of the top contributors for the past three years.
All variant submissions shared with ClinVar are anonymous because they cannot be linked back to a particular person. The information that Invitae shares adheres to current patient privacy standards and a strict privacy policy for patients’ protection and peace of mind.
In addition to stringent HIPAA privacy and security regulations, labs have to meet or exceed the standards set by the Clinical Laboratory Improvements Amendment (CLIA). The College of American Pathologists (CAP) also sets strict benchmarks that labs must meet to receive accreditation (official recognition).
“Broadly sharing variant data is critical to advancing genetics and the science of precision medicine. We honor patients by contributing variant data while protecting their privacy and leveraging test data to benefit us all.” — Deven McGraw, Lead, Data Stewardship and Data Sharing at Invitae
Invitae recently exceeded 1 million total variant submissions to ClinVar, making it the top contributor of all time. The ClinVar team now plans to scale to support five times Invitae’s previous contributions, meaning more genetic answers for patients in the future.
Data sharing through ClinVar is unique in that it allows ongoing:
- inter-laboratory quality control
- detailed peer review of variant classifications
- consensus classification by the global community of experts
No other mechanism, including published scientific papers, solves these crucial problems. By contributing to ClinVar, Invitae is helping to expand access to and communication about the relationship between genetic variants and observed health status. As time goes on, more labs are starting to understand the importance of data sharing.
The technology behind solving gene variants
In addition to supporting community efforts for improved variant classifications, contributing to ClinVar helps develop and validate new technology in academic and research settings.
Invitae also participates in this research, leveraging definitive variant classifications in ClinVar to develop and validate new evidence for variant interpretation. These include machine learning models and large-scale assays designed to determine the functional impact of clinically observed variants.
Invitae has established a refined framework for accurately determining variant classification through its Sherloc technology. The power of Sherloc is that it’s designed to provide a system for clinical geneticists to combine an expanding array of variant evidence into a reproducible classification that meets or exceeds guidelines set forth by the American College of Medical Genetics (ACMG) and the Association for Molecular Pathology (AMP).¹
Often, there isn’t enough evidence for each variant to reach a definitive classification. To help address this challenge, Invitae has developed an evidence modeling platform to develop and validate sophisticated methods that learn information about genes and proteins to better estimate the impact of sequence variants on disease and phenotype.
The aggregated data on ClinVar are essential in the development of these models. They serve as one method for labeled training, prospective evaluation, calibration and clinical validation of new models. If the models perform well, they are translated within the Sherloc framework, providing additional evidence for variant classification.
When Invitae submits variants with supporting evidence to ClinVar, they’re reviewed and scrutinized by the community of clinical and molecular geneticists worldwide. These experts compare the submitted evidence with other labs’ interpretations of the same variant.
There is no gold standard for quality control in variant interpretation. The closest we have now is community consensus from professional experts with deep knowledge of the rules and guidelines for variant interpretation.
“It’s not just that different labs may interpret the evidence about a variant differently. More commonly, they can have different amounts of non-public information about a patient’s clinical presentation or family history or may have previous experience with that variant in other patients, affecting how they interpret a variant. Breaking variant interpretations out of their silos in clinical labs and sharing them provides an important tool for improving patient care.” — Dr. Robert Nussbaum, original co-investigator on the NHGRI/NIH grant that established ClinVar, and Chief Medical Officer at Invitae
Why ethical data sharing is vital to patient healthcare
Behind every genetic variant is a patient seeking answers that could impact their health and their family’s. The more variants we classify and share widely, the greater the odds of developing new therapies to address them and identifying people who could benefit.
The American College of Medical Genetics (ACMG) recently stated that sharing de-identified genetic data is “crucial” to advancing personalized medicine. The ACMG joins the American Medical Association (AMA), the National Society of Genetic Counselors (NSGC) and other professional societies in making this strong practice recommendation.² ³
Not only is it essential that labs ensure their results are accurate, but they must also help other labs do the same. While that’s not yet the industry standard, Invitae is determined to set the benchmark, leading by example on how to do so ethically.
Questions to ask your genetic testing lab
If you don’t test with Invitae, we encourage you to ask your lab the following questions. Find out how they help advance the field of genetics (and better patient care as a result) and what measures they take to protect patient privacy.
Do you have a privacy policy, and is genetic information de-identified?
Reviewing a lab’s privacy policy before submitting sensitive data is crucial.
Do you share all clinically observed variants and their classifications?
Some labs limit sharing only to variants with “high confidence” classifications.
Do you provide the available data to a public database?
Many databases do not meet the same open, free, unrestricted, public access standards as ClinVar, so some labs are not submitting data to public databases that will increase access.
How often do you share data and update previous submissions?
Regularly submitting new data and updates to existing data helps share the most current information with the global community.
To learn more about Invitae’s expert-driven, AI-enabled approach to genetic testing, read more on Invitae Generation™ here.
References
- Richards S, et al. Genetics in Medicine. 2015;17(5):405–424.
- ACMG Board of Directors. Genetics in Medicine. 2017;19(7):721–722.
- Clinical data sharing. National Society of Genetic Counselors. Published November 25, 2020. Accessed December 19, 2022. https://www.nsgc.org/Policy-Research-and-Publications/Position-Statements/Position-Statements/Post/clinical-data-sharing.
