PGx Analysis in VarSeq: A User’s Perspective

Editor's Notes

1. Casey’s intro
2. Casey reminding them to input questions
3. Darby Intro
4. So with the recent release of VarSeqs pharmacogenomics capability, Solomon and I have gotten some feedback from our users on workflow topics when setting up your PGx analysis in our software. We wanted to provide some more exposure to key components of the PGx pipeline that will assist you when moving through the validation process.
5. Before we start diving into the subject, I wanted mention our appreciation for our grant funding from NIH. The research reported in this publication was supported by the National institute of general medical sciences of the national institutes of health under the listed awards. We are also grateful to have received local grant funding from the state of Montana. Our PI is Dr. Andreas Scherer who is also the CEO at Golden Helix and the content described today is the responsibility of the authors and does not officially represent the views of the NIH. So with that covered, Before diving into today's topic, I'd like to offer some background and context on what Golden Helix brings to the table as a company
6. Golden Helix as a company has been around over 25 years now and established itself as a premier tertiary platform for both next gene sequencing and array data. Our primary mission to develop high quality general purpose bioinformatic software with a key focus on facilitating routine clinical applications. Our software solutions are highly scalable for everything from panels to genomes but are highly automatable when processing a large number of samples in your lab. This combined with our subscription based business model, allows users to process as many samples as needed, without us chipping away at your revenue stream with most per sample applications on the market. The assays designed in our software are built by the user and span a wide spectrum of uses, including somatic workflow, hereditary cancer, rare diseases, prenatal testing and family based analysis, and of course our recent addition of Pharmacogenomics. Our customer base is also wide spread with users in government and testing labs, hospitals, universities, and many research and pharmaceutical labs.
7. Lets review where our tools fit into the bigger picture of an NGS pipeline. The Golden Helix software is in tertiary analysis stage following output from your preferred secondary analysis solution. You can see here we have a partnership with Sentieon to provide the secondary solution for labs who may need it, but overall, our VarSeq software is agnostic to the variant caller being used. Moreover, we are also agnostic to the sequencer and have had a number of demonstrations on VarSeqs’ capability of processing long read data as well. With VarSeq, users will be able to import any scale of data from panels to genomes, for both short and long read data, from any sequencer to quickly annotate and filter variants to evaluate and report on the clinically relevant findings. And recently we’ve announced our medical device certification of our VarSeq software suite
8. For European labs or any lab processing European samples, we’re happy to announce our recent ISO13485 certification as the result of our diligence in creating a thorough quality management system coupled with our manufacturing process. In the near future, labs developing in house tests will need to adhere to the IVD regulations, and can rely on streamlining integration of our software having the certification already in place. When implementing our tools, our support staff is also ready to guide you through our own user onboarding and certification process bringing your users up to speed as we move through the workflow validation process together. Now let’s get to todays topic with Pharmacogenomics
9. If you are a regular at our webcasts, you may have seen the recent presentations on the release of the PGx capabilities in VarSeq. Th
10. I encourage you to go back and watch the earlier webcasts as they go into deeper detail on the nature of our PGx solution. The star alleles we’ve integrated thus far are based on the Clinical Pharmacogenomic Implementation Consortium or CPIC database. CPIC is an international consortium dedicated to facilitating the use of pharmacogenetic tests for patient care. Their goal is to create clinical best-practice guidelines for pharmacogenomic testing so that clinicians, health care providers and vendors can leverage genetic laboratory results into actionable prescribing recommendations. CPIC provides this publicly available database that links phenotypes to specific allele combinations and provides prescribing recommendations based on this phenotypic information, with each recommendation being ranked based on the level of supporting evidence. This database has been indispensable in the development of our own pharmacogenomic solution.
11. Let's take a moment to break down the inputs to the PGx algorithm. As Darby just elucidated, star alleles are at the core of our PGx workflow, which makes pharmacogenomics unique in the NGS space in that we're just as concerned with reference calls as we are with variant calls. This requires special treatment during the secondary analysis phase. Ultimately, our goal is to produce a VCF containing both variant and reference calls, or calls with a 0/0 genotype, for specific loci defined by the CPIC database. As Darby mentioned, flexibility and the ability to easily integrate with existing tools are priorities when our team develops new workflows in VarSeq. Hence, we have a few different solutions for getting NGS data into VarSeq. We have developed a Sentieon pipeline to implement forced calling of CPIC variants that is available to our Sentieon customers, but we also provide support for gVCF inputs, array data conversion, and other secondary analysis pipelines capable of implementing forced calls.  Let's dissect the two routes for producing PGx input with NGS data. The first is using a given VCF with must-call variants, derived from the CPIC database. We have implemented this workflow with Sentieon, but it can just as easily be replicated in other secondary analysis tools. Fundamentally, this just means performing standard alignment and variant-calling steps and providing a list of variants where the caller must evaluate a variant or reference call. We'll take a deeper look at this during our demonstration of the software in just a bit. Alternatively, we can handle the reference call inference during the VarSeq import step with a gVCF as input. A gVCF contains a list of intervals where no variants were detected, and hence can be used to infer reference calls that overlap with these intervals.  In either case, both variant and reference calls are subject to the coverage of your NGS sample. We can only make calls where there is coverage, and this is reflected in the PGx diplotype calling algorithm as well as the report, which Darby will show off shortly. This means that you can be confident the PGx workflow is only working with the genes you are including in your assay. This is also true for array data, which we're happy to support as well. 
12. Array data is still common for PGx testing. We can handle this data in VarSeq by converting from a text file source into a VCF and importing into VarSeq. Just like with NGS data, this will ensure that the PGx analysis is subject only to the variants of interest for your workflow. We've seen a plethora of text outputs from PGx array data, but a common theme is a list of RSIDs and associated alleles. We can cross-reference these with our PGx must-call VCF in a conversion script and produce something similar to what the NGS approach would yield. So, regardless of where your data is coming from, we have a solution to get it into VarSeq and run it through the PGx diplotype caller. I'm going to hand things back to Darby here to show off what that process looks like.
13. Thank you Solomon for providing a simple overview of the genotyping options that are supported. As Solomon had mentioned, if you are working with array data, we’ve had numerous iterations of helping our users convert this data to VCF so to import into VarSeq. Now let’s review what happens next once the data is in the software.
14. After running the PGx Variant Detection and Recommendations algorithm, a clinical report can be generated using VarSeq’s customizable reporting system. Clinical reports are generated using an easy-to-modify Microsoft Word report template and VarSeq comes shipped with an initial PGx report template that serves as an excellent starting point for creating custom reports. Information included in this report includes Implications for Current Patient Medications, Gene-Drug Interactions, Prescribing Recommendations, Phenotype Associations, and a description of all tested alleles. The calling of diplotypes, annotation of phenotypes, and reporting of recommendations is performed by VS-PGx in just a few simple steps with minimal user involvement. While the annotations and report templates can be customized, the annotation tracks and report templates provided by VarSeq have everything you need to start annotating and reporting all alleles defined in the CPIC database.
15. After running the PGx Variant Detection and Recommendations algorithm, a clinical report can be generated using VarSeq’s customizable reporting system. Clinical reports are generated using an easy-to-modify Microsoft Word report template and VarSeq comes shipped with an initial PGx report template that serves as an excellent starting point for creating custom reports. Information included in this report includes Implications for Current Patient Medications, Gene-Drug Interactions, Prescribing Recommendations, Phenotype Associations, and a description of all tested alleles. The calling of diplotypes, annotation of phenotypes, and reporting of recommendations is performed by VS-PGx in just a few simple steps with minimal user involvement. While the annotations and report templates can be customized, the annotation tracks and report templates provided by VarSeq have everything you need to start annotating and reporting all alleles defined in the CPIC database.
16. After running the PGx Variant Detection and Recommendations algorithm, a clinical report can be generated using VarSeq’s customizable reporting system. Clinical reports are generated using an easy-to-modify Microsoft Word report template and VarSeq comes shipped with an initial PGx report template that serves as an excellent starting point for creating custom reports. Information included in this report includes Implications for Current Patient Medications, Gene-Drug Interactions, Prescribing Recommendations, Phenotype Associations, and a description of all tested alleles. The calling of diplotypes, annotation of phenotypes, and reporting of recommendations is performed by VS-PGx in just a few simple steps with minimal user involvement. While the annotations and report templates can be customized, the annotation tracks and report templates provided by VarSeq have everything you need to start annotating and reporting all alleles defined in the CPIC database.
17. Before we start diving into the subject, I wanted mention our appreciation for our grant funding from NIH. The research reported in this publication was supported by the National institute of general medical sciences of the national institutes of health under the listed awards. We are also grateful to have received local grant funding from the state of Montana. Our PI is Dr. Andreas Scherer who is also the CEO at Golden Helix and the content described today is the responsibility of the authors and does not officially represent the views of the NIH. So with that covered, Before diving into today's topic, I'd like to offer some background and context on what Golden Helix brings to the table as a company