Genomics in the Cloud

Chapter 6. GATK Best Practices for Germline Short Variant Discovery

By now, we’ve introduced all of the major concepts and tools, and we’ve walked through a few examples, calling variants with the GATK HaplotypeCaller and exploring its output. It’s time to turn up the dial on the analytical complexity and take a closer look at the three GATK Best Practices workflows.

In this chapter, we explore the most widely used variant discovery pipeline in the genomics community: the GATK Best Practices for germline short variant discovery in multiple samples, which is designed to take advantage of the benefits of including more data in the analysis process. We begin with an overview of the initial data processing and quality control steps that are necessary to render the data suitable for analysis. Then, we’ll dive into the details of the joint calling workflow. In the final part of this chapter, we also cover the highlights of an alternate workflow that supports single-sample calling for use when multisample calling is not an option, as in the context of clinical analysis.

Data Preprocessing

As we discussed in the genomics primer, the data produced by the sequencing process is not immediately suitable for performing variant discovery analysis. We need to preprocess that raw output to transform it into the appropriate format and clean up some technical issues. In this section, we outline the key steps and their rationale. However, we don’t run through the entire preprocessing workflow in ...