How Invitae detects challenging genotypes

Dianalee McKnight, PhD, FACMG

Modern genetic sequencing is driven by a fast and cost-effective DNA technology called next-generation sequencing (NGS). This powerful tool is used by many clinical labs to give timely answers to patients with medical genetic concerns. But there are some facets of our genome—like pseudogenes, repetitive sequences, and copy number variants—that trip up standard NGS. We call these challenging variants “hard-to-do’s” and our data show that as many as 14% of pathogenic or likely pathogenic variants fall in this category. While some labs turn to costly and cumbersome tests to assess hard-to-do’s, Invitae has carefully customized the chemistries and computational analyses of our NGS to bring even the most challenging variants into the fold. In some cases, our NGS-methods capture even more clinical information about hard-to-do’s than other techniques. 

Not all NGS approaches are the same. The technology is complex and there can be many differences in the details, such as the depth of sequence, or how many times each spot in the genome is read by the NGS machine. Invitae sequences all samples to a high depth to ensure we always glean ample information about each genetic locus. We follow this with thoughtful analysis of sequencing data with both public and internally developed algorithms. Critically, we have made great friends with robots. Our highly automated lab workflow provides consistent replication from one sample to the next at a level of precision that a human can’t reach. By combining high sequencing depth, great uniformity of sample processing, and smart, integrated bioinformatics, we unlock new capabilities of NGS. 

One achievement leverages our highly uniform NGS workflow to identify small copy number variants (CNVs), such as single-exon duplication or deletion. While these hard-to-do’s were well known for certain genes associated with disease like BRCA1 or DMD, they were largely underappreciated in the rest of the genome because the techniques required to evaluate them were too burdensome for universal use. Building on our highly automated and uniform sample processing, Invitae was able to develop algorithms that can identify intragenic CNVs without additional molecular techniques. Once we opened the floodgates with our pioneering CNV detection methods, we could see that CNVs in almost 10% of individuals with a pathogenic finding. Many of these would be missed with other systems.

We have also used NGS to address specific genes with challenging genotypes. These include Lynch syndrome gene PMS2 and spinal muscular atrophy gene SMN1. Both have near identical pseudogene copies that could not be differentiated with generic NGS methods. Traditionally, labs have turned to time-consuming and expensive assays beyond NGS that lead to higher prices and delayed results. To bring these hard-to-do’s into our comprehensive, single sample workflow, Invitae developed NGS-compatible methods that can detect variants in both of these challenging genes with very high accuracy and sensitivity. What’s more, for both conditions, our methods provide more information than traditional approaches—an additional 25% more clinical sensitivity for PMS2 and more information about SMN1’s near-identical copy, which can affect prognosis. 

We have tackled many other hard-to-do’s, like changes affecting the triplet repeats in FMR1 that are associated with fragile X syndrome, the recurrent inversion in cancer-linked MSH2, and other genes with nearly identical copies like the HBA1/HBA2 complex associated with alpha thalassemias. Each of these special attention cases fits into our high-throughput, economical workflow. This enables testing for a broad variety of genotypes at greater clinical sensitivity from a single patient sample without added cost. 

Technically challenging genotypes are not rare.  In a study of almost half a million people, we found that the hard-to-do’s we can already detect represent 14% of pathogenic or likely pathogenic reports. We are continuing to add more of these challenging, special cases to our NGS workflow, building on our already deep investment in our platform so we can deliver genetic insights at a scale previously considered out of reach.

You can learn more about Invitae’s approach to detecting challenging genotypes in a recent webinar hosted by me and my colleague Dr. Swaroop Aradhya.