Webinar: Accelerating Patient Diagnosis with Automated Literature Curation for Inherited Disease
Dr. Mark Kiel and Birgit Funke discuss an approach for automating variant classification based on the American College of Medical Genetics and Genomics guidelines.
This webinar discusses an approach for automating variant classification based on the American College of Medical Genetics and Genomics guidelines.
Variant interpretation presents a bottleneck for many labs, posing a challenge to the broader adoption of precision medicine. The ACMG/AMP variant classification framework has provided a foundation for this process, but several key obstacles remain. In particular, the identification and prioritization of key publications presents a time- and resource-intensive challenge for many labs.
Watch this webinar for a discussion on how Veritas Genetics collaborated with Genomenon to develop a new Literature Classification Tool within the Mastermind Genomic Search Engine to address these challenges.
Genomenon CSO and co-founder Dr. Mark Kiel and Veritas Genetics VP of Clinical Affairs Birgit Funke discuss the accelerated literature curation process, including:
- How to more efficiently identify and prioritize publications by ACMG variant classification guidelines.
- How increased specificity and immediate access to annotated search results accelerates variant interpretation workflow.
- How increased sensitivity in literature search results in fewer false negatives.
The speakers discuss promising Veritas trial results that point to a better and faster variant interpretation workflow using the new Literature Curation tool in the Mastermind Genomic Search Engine.
Webinar Q & A
If Mastermind is Missing an article or variant, how would a user go about reporting this to Mastermind?
If a reference is found to be missing or otherwise variant or gene or disease data is not identified in Mastermind, please reach out to us through the Contact Us feature within Mastermind or otherwise email us at email@example.com.
One of the most important needs for a clinical lab is the confidence that a null search really means that there are no publications. Can you speak to this a little more and point our the degree of confidence when no publications are identified in Mastermind?
No search engine will ever have 100% sensitivity. We believe that Mastermind is much more sensitive than any other method or variant database software available. We have done promising benchmarks in the past (see our blog) and are assembling data now that speaks to the statistical confidence users can expect of a null result when using Mastermind.
Is there any integration with OMIM?
Mastermind does not yet have links into OMIM nor does OMIM have links into Mastermind. We hope to engage these conversations soon.
Is Mastermind integrated into some existing Variant interpreter software, eg Illumina BaseSpace Variant Interpreter, and Agilent Alissa?
We are working on several third party software integrations at present.
Would you consider making a track available on UCSC as to whether Genomenon has identified literature about a gene or variant within a gene? This could be super-helpful.
We’ve just begun having conversations with the data scientists at UCSC to provide a data track of all of the Mastermind variants in the UCSC browser.
Could you mention how Mastermind will index and incorporate rsIDs in the future?
Mastermind already indexes the literature for rsIDs. This variant type can be searched in the user interface as well.
Is every full-text paper indexed by Mastermind available to read (fully) to a Mastermind user directly, or are some/many of the full-texts found by Mastermind only accessible behind a publisher’s paywall?
The indexed content that is visible to a user through the Mastermind interface (i.e. title, highlighted abstract and full-text sentence fragments) is available to all users. To access full-text PDFs, a user must already have access to the content – whether it is an Open Access article available to all or otherwise the Mastermind user has a subscription to that publisher’s content.
Is Mastermind cognizant of LOVDs or other databases and their conclusions regarding variants?
Mastermind currently links out to some third-party databases, and will be expanding this in the future.
Is the literature search real-time?
No. The indexing process takes place continually on the many thousands of newly published references, and the data in the Mastermind database is updated weekly. We regularly re-index the existing literature to ensure it is up to date.
Do you have any plans to take further steps to automate ACMG guidelines based classification based on literature data, such as provide preliminary variant classification for curator to review?
Yes, we definitely do. Not the least of which is working with Birgit to continue to evolve and improve the specificity of the search results. We see this as an important tool because there are too few individuals who are currently trained to do variant interpretation. Our approach with building the software is meant to facilitate the efficiency of that curation work. We are hoping to build more automated ACMG classification features and specifications in the Mastermind software to provide users with a more efficient and painless manual review.
Does your search include supplemental materials?
We are adding supplemental materials into our indexing process this quarter.
How does Mastermind search compare to Google Scholar search in terms of gene disease level search?
Favorably. You can read the results of our initial benchmark in our blog here.
Does Mastermind support variant searches for dups/dels including large contiguous gene events?
Yes. Mastermind affords the user the ability to search for del/dup variants if they are described using appropriate HGVS nomenclature.