Optimizing sample throughput while maintaining quality control is, arguably, the crux of any profitable lab performing a high volume of analysis. We’ve discussed in great detail the flexibility of VSPipeline, our command line automation tool, as well as our full-stack FASTQ to report automation capabilities, and today I’d like to further elucidate on how QC methods fit into the picture.
The ability to seamlessly investigate and reanalyze projects created with our automation tools is a core component of our design philosophy at Golden Helix. Exporting QC metrics from each step of the pipeline is a great way to quickly assess the overall quality of a sample as well as precisely pinpoint which component of the pipeline needs to be manually assessed in the event of QC failure.
Let’s break down a hypothetical FASTQ to report pipeline and explore the possible outputs. Starting with the secondary analysis pipeline, the generation of abundant sequencing, alignment, and variant-calling metrics is common to most pipelines. These metrics can be stored, summarized, and propagated through the rest of the pipeline, and issues related to secondary or primary analysis can then be easily processed.
Next, let’s consider a few metrics that could be exported from a tertiary analysis pipeline in VarSeq. An obvious high-level metric that can be output from this stage of the overall automated workflow is granular coverage statistics calculated in VarSeq for the sample, which can be used to assess the sample’s overall quality as well as per-region and per-variant viability. Beyond overall coverage statistics, high-level statistics like the number of filtered variants or variants meeting predefined criteria can be easily exported and summarized.
The next step in a full-stack automated workflow is to import variants into VSClinical, VarSeq’s clinical analysis tool. As we’ve seen, VSClinical is extremely flexible, and arbitrary file types can be exported summarizing the variants imported and any matched interpretations. Also housed within VSClinical is the last and most obvious piece of QC information, which is the report itself. Any of the above-mentioned metrics can be propagated through the pipeline and included in the clinical report, which can be used as an at-a-glance QC indicator.
In summary, all of these outputs tell a detailed QC story about a sample or set of samples passed through an automated clinical pipeline. The main takeaway should perhaps be the flexibility of the methodology described; at each step, the inputs and outputs can be defined to meet a lab’s needs and maximize efficiency. If you’re curious about any of these options or want to learn more about any of our tools, please don’t hesitate to reach out to [email protected]. We look forward to supporting your NGS needs!