The Wide Spectrum of Next-Generation Sequencing Assays with VarSeq

Editor's Notes

• #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, lets take just a few minutes to talk a little bit about our company Golden Helix.
• #6: Golden Helix is a global bioinformatics software and analytics company that enables research and clinical practices to analyze large genomic datasets. We were originally founded in 1998 based off pharmacogenomics work performed at GlaxoSmithKline, who is still a primary investor in our company.  VarSeq is our flagship product and serves as a clinical tertiary analysis tool. At its core, it is a variant annotation and filtration engine. But also gives users ability to automate the AMP or ACMG variant guidelines. VarSeq also has the capability to detect copy number variations scaling from single exome to large aneuploidy events. Lastly, the finalization of variant interpretation and classification is further optimized with the VarSeq clinical reporting capability. Users can integrate all of these features into a standardized workflow, save as a template, and automated the major of this tertiary process with our Vspipeline tool Paired with VarSeq is VSWarehouse which serves as a locally installed repository for the large amount of useful genomic data collected from your samples. Warehouse not only solves the issue of private data storage for ever-increasing genomic content, but also is fully queryable and auditable and allows for the definability of user access for project managers or collaborators. Lastly, our research platform, SVS, enables researchers to perform complex analysis and visualizations on genomic and phenotypic data. SVS has a range of tools to perform GWAW, genomic prediction, and RNA-Seq analysis, among other common research applications.
• #7: Over the course of 20 years, Our software has been very well received by the industry. We have been cited in thousands of peer-reviewed publications in reputable journals, as one testament to our customer base.
• #8: We work with over 400 organizations all over the globe. This includes top-tier institutions, like Stanford and yale, government organizations like the NCI and NIH,  clinics such as Sick Kids, and many other genetic testing labs.  We now have well over 20,000 installs of our products and with 1,000’s of unique users.  So how is this relevant to you?
• #9: At Golden Helix, we strive to be the complete solution for our customers which is defined across seven pillars comprising the Golden Helix difference. Golden Helix offers a single software solution that encompasses germline, somatic, and CNV analysis. Our software is also highly scalable, supporting gene panel to whole genome sequencing workflows. With our complete automation capabilities, we now offer a FASTQ or VCF to report pipeline. Our software is locally deployed, or installed in cloud, respecting our customers data privacy Our business model of annual subscription per user means you are able to increase your workload without increasing analysis fees. And the license comes with unlimited training and support from our FAS team to facilitate Customer Success. 
• #10: So now that we’ve talked a bit about the value of Golden Helix as a company, lets switch to the topic of todays discussion and cover the versatility of NGS workflow design with VarSeq JEN: Thank you Darby, I’m happy to be here!
• #11: J: Before we head into a few project demonstrations, let’s discuss the logic surrounding VarSeq and VSClinical. VarSeq and VSClinical leverage ACMG and AMP guidelines to determine pathogenicity or oncogenicity for both small variants and copy number variants. There is a lot to consider when reviewing all the components for the ACMG or AMP guidelines manually, which makes all the hard work in automating this process so valuable. In its most basic form, VarSeq can be seen as a three stage process. We have Stage one, which encompasses the importing, detection, and filtering of variants from your VCF file. All the imported or detected variants pass through a user define template that is based on a variety of public databases and algorithms. ​Today we will see how this is applicable to small gene panels all the way to whole genome projects. ​  After isolating our germline SNVs, INDELs, or CNVs of interest, these will be carried into stage 2 of the analysis, where they will be analyzed in VSClinical through a review of the ACMG or AMP guidelines. ​ ​ Following the variant evaluation, we are able to render the final interpretation results in a complete clinical report which is the third and final stage of analysis. The overarching steps described here can now be applied to any of the workflows we will be getting into, and these will range in complexity from a very small gene panel to whole genome workflows. So let’s take a look at how a panel is traditionally constructed, along with the upsides and downsides to this process. **NEXT SLIDE**
• #12: Most of us here are familiar with the process behind constructing a sequencing kit for small panels, if not still using this method. This method does have some advantages, for example, companies make the design of the primer kits simple and straightforward. Additionally, this route of analysis is fairly efficient when looking at simple panels, maybe only a few genes, especially in a situation where a lab has high sample turn over for a limited number of panels. Let’s take a look at this process. First, I know which genes I want ahead of time, and have submitted those to the primer company. After receiving my kit, I will sequence my samples, then I will go through the secondary analysis, converting my FASTQs to BAMs and VCFs. After that is the matter of applying a simple filtering strategy to isolate my variants of interest, and then reporting on my clinical findings. Our targeted NGS panel is going to lead to highly specific results, and the workflow template can be applied to this panel over and over again. Here we are going to use a BRCA1, BRCA2 panel, resulting in a loss of function variant in BRCA2. Our path to reporting is expedited, as we can supplement our report with prerendered interpretations from our CancerKB database. Now, the problems here become obvious when we look at changes in the market demands. **CLICK** For example, the need to update the gene panel for a more extensive report. Here, that would be if I wanted to report on additional genes such as ATM, CHEK2, and PALB2 in addition to BRCA1 and BRCA2. Changing the panel at this point means the clinician will have to go back to version one, redesigning and resubmitting a gene list to a company. With this method, you are limited to the scope of the panel for genomic findings. If we want to increase our flexibility for reporting, it would make sense to begin with sequencing the exome if not genome to maximize potential on diagnostic yield for ever expanding panels and apply them virtually rather than go through a limited sequencing kit.
• #13: A good example here would be transitioning to an exome level data set, where you can redesign your panels easily by applying virtual panels. With this model, we would version those panels for use in audits. Applying virtual panels can save time in the long run as there is no need to design specific panels prior to sequencing. The ability to retroactively apply phenotypic searches or gene panels to Whole exome sequencing level data gives an enormous reengineering potential that can lead to specific insurance reimbursements. Moreover, when it comes to reporting, VSClinical has a number of predesigned report templates that can easily be applied to numerous assay designs. The ability to automate this process, going from FASTQ to reporting with VSPipeline can also expedite sample turn over in the lab. Now, this panel construction can seem daunting, but there are guidelines from the ACMG group that can take you through this process. Here, we are looking at several examples of broad disorder panels with the number of genes in each panel, along with more specific subpanels and their respective number of genes. I also have a link to the ACMG guidelines paper that gives details for these panel constructions. The idea is, by starting at whole exome sequencing, we do not have to start at one kit, for example, dilated cardiomyopathy, and then start over with a new kit if the phenotype is revised to hyper-tro-phic cardiomyopathy. I simply have to apply the new virtual panel to the data set. VarSeq has the tools to make this process simple, including the Gene Panel Manager that we will take a look at in our demonstration. These tools make constructing panels as easy as possible. Overall, the pros with applying virtual panels include the ability to expand on diagnostic yield, while remaining adaptive for all potential reimbursement and reporting for evolving panels. The cons are, obviously, the construction of the panels themselves, but there are tools and guidelines to help you in this process. The ease of use of panels to reporting is best exemplified with our support of TSO500 somatic analysis.
• #14: In addition to the construction of virtual gene panels, I wanted to highlight to our viewers, the streamlined integration of our TSO500 workflows with comprehensive genomic profiling. Starting from the sequencing stage, VarSeq analysis is agnostic as to the origin of the somatic data, as long as that data is in a standard format. This means we can cover solid tumor sequencing, liquid biopsies, FFPE tissue samples, and more all under the same workflow umbrella. The variants supported for analysis include SNV, CNVs, structural variants, and fusions. The workflow, starting with panels, then virtual panels, and now our TSO500 somatic panel, has been getting gradually more involved. As we become tumor type specific, we are going to bring in more cancer elements, such as specific cancer sub-panels, or cancer annotation sources like CIViC and DrugBank. Moving into VSClinical, you will see that our comprehensive genomic profiling will bring in additional fields, like genomic signatures, TMB, MSI, or HRD. We will bring in negative findings that are specific to tumor types. We will also bring in immunotherapy biomarkers and relevant clinical trials. This comprehensive overview is enhanced by our inhouse Cancer Knowledgebase. CancerKB has prerendered clinical interpretations scoped for TSO500, and is ready to supplement biomarker reporting. Last, we have our automation component, where with VSPipeline, we can go all the way from FASTQ to clinical report, ready for a lab director sign off. The large amount of information available truly makes the TSO500 somatic panel with VarSeq the comprehensive somatic analysis. Now that we have covered our somatic panels, we are going to pivot, and I am going to hand things back to Darby to tell us about working with whole exome, germline workflows. FFPE- formalin fixed paraffin embedded tissue samples CGP- comprehensive genomic profiling TMB, MSI, HRD
• #15: Thank you Jen for breaking down the evolution of the gene panel and what tools we have to simplify somatic analysis. Switching to the germline topic, an example of a typical exome analysis would be Trio or family based analysis, for two parents and any number of probands. From a workflow perspective, this is relatively sophisticated when designing filters that account for different inheritance models. The inheritance model strategy can also be utilized with VarSeq called CNVs as well. One common hurdle with large scale data, like the exome in a clinical space, is that each patient brings with them a unique phenotype or diagnosis. Because of this, panels can be difficult to construct for each unique case. The solution is implementing GHI’s PhoRank algorithm that allows users to prioritize variants that are in genes highly associated with the diagnosed phenotype. This helps users overcome the required adaptation to running larger scale data. When doing the variant evaluation, VSClinical accounts for inheritance based ACMG criteria for the clinically relevant variants. Just as is the case with panels, this process is highly automatable. One drawback to consider with ‘trios’ is the increased costs for each sample of the family, and increased analytical time. It is for this reason that we wanted to show you how you can reach the same outcome with a singleton analysis rather than run the trio.
• #16: So the outcome is potentially similar to that of the trio analysis but the upside for the single sample workflow is efficiency, reduced cost and overall simplicity. The filter chain will be a bit different as we let go of the possible inheritance models, but you’ll see in the demo there are a number of creative ways to still reduce the number of variants you need to evaluate. One thing to consider is the con of switching to single sample from trios, you lose the ability to run any additional risk analysis on the parents if they seek to have other offspring. However, as you saw in the slide before, if the risk analysis is needed, the trio template is available and ready to be applied for our users workflows. Additionally, you’ll see in our trio example the context of variants that are in the cis and trans state can affect relevant classification criteria and one thing that will assist here in the near future is the phased genotypes with long read technology if limiting your analysis to just the proband. This is all in the context of the exome, but we still might consider running the genome as it provides the ultimate diagnostic yield.
• #17: Here is an example of the coverage from a genome on top compared to the targeted exome coverage just underneath. The depth of coverage for the exome looks quite nice and any allele in the exon or perhaps just on the edge near a canonical splice site would be identifiable. However there are severe limits, and I created an example to purposefully demonstrate the inability for the exome to capture a deep intronic pathogenic allele that no clinical lab would want to miss. This is the diagnostic power and value of the genome, as it will have the adequate coverage necessary to detect alleles like our example MHS2 novel splice variant that is well known pathogenic. So obviously genomes come at a cost for computation burden but our tools support the transition to genomes as they become more common due to reduced cost and want to assess population level genetic findings for various populations. Not only do our users have the ability to streamline the process of genomes with tools like VS pipeline on your server but we also support CRAM files now to tackle the data storage issue that comes with reviewing coverage at a genome level. Now that we’ve covered some workflow and data scale highlights, lets review our example variants for the software demonstration
• #18: So we have covered a lot of great highlights concerning the evolution of NGS assays and our demonstration is going to give examples moving from the panel to genome level analysis. X Y Z And now onto our product demonstration~
• #20: In summary, we are trying to facilitate the evolution of NGS assays with the goal of standardizing the exome and the genome. We do this while supporting the different variant types such as SNVs, INDELs, CNVs, and fusions. When panels fall short, users can rely on a phenotypic based variant prioritization with the PhoRank algorithm. Additionally, we aim to expedite the somatic review with our AMP evaluation tool in VSClinical paired with CancerKB, with a minimal standard held against the TSO500 gene list. In any scenario, be it panels, exome, genome, germline or somatic, all workflows are highly automatable, to save the user time and cost. With assay design in Golden Helix software, comes limitless training with our FAS team.
• #21: 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, lets take just a few minutes to talk a little bit about our company Golden Helix.
• #23: Title: Project resources Category: Data, Images Tags: table, project, resources, data, cost, picture, image, vial, bottle
• #24: Again, I want to mention how grateful we are we are thankful of grants such as this which support the advancement and development of our software to create the high quality software you'll see today. So with that covered, lets take a few minutes to talk a little bit about our company Golden Helix.