Where is the future headed?
With an ever-increasing number of options for diagnostic sequencing, the decision of whether to do a gene panel, exome, or whole genome becomes complicated.
I’ve been debating this a lot, both on social media and with work colleagues recently. There’s lots of great discussions and papers around exome v whole genome sequencing v gene panel. Why not join the debate and give me a tweet at @diagnosticsjoe?
— BarbaraJennings (@GeneticsMBBS) January 23, 2016
Before you decide on the type of sequencing, for the uninitiated it’s worth pointing out the various library prep methods currently available and what the consequences for downstream applications. What library preparation method are you going to use? Hybridization? Amplicon? For your choices, there are the sonication + purification via magentic beads as seen in SureSelect and SeqCap or perhaps more cutting-edge (defined by me in this respect as requiring low DNA/RNA input), you can opt for one of the amplicon methods. It’s worth noting that even within specific prep methods there’s even further differences. eg. HaloPlex, fragments genomic DNA with restriction enzymes, then uses probes complimentary to the 5′- and 3′-ends of each fragment for the creation of a targeted amplicon; this is followed by PCR primer annealing and PCR amplification (diagram below).
Those in favour of exome sequencing (use #exomesrule):
The main argument FOR is certainly the certainty of diagnosis. Exome sequencing is suited best to diseases (conditions) that have multiple phenotypes that are expressed clinically BUT do not have a clear diagnosis, such as neurodevelopmental and other neurological presentations. In converse: panel testing is best suited to patients with a clear diagnosis and for which panel testing yields a reasonable detection rate, such as cardiomyopathy, retinal disease, and hearing loss.
In favour of panel sequencing (use #panelsrule):
Again, perhaps the biggest argument is from your wallet (!) But yet perhaps the largest reason for using panel sequencing and NOT exome is that whole-genome or exome sequencing is that those techniques don’t always capture problematic regions or the regions you’re actually interested in. Targeted panels can be enhanced in such regions, such as: Indels, GC-rich regions, paralogous genes, and repetitive regions.
What to do with VOUS?
That is the question…The cost of clinical-grade WES is high, typically > US$4500 (higher when done as a child–parents trio), with a long turnaround time of three months on average. These drawbacks are primarily driven by the interpretation challenge. WES typically uncovers numerous variants and identifying the one causal variant can be a challenge. Despite the development of several tools that aid in the automation of WES variants, the requirement for manual inspection and expert analysis remains.
I’ve briefly compared some of the key features of panels v WES in the table below:
|Size||50-100 genes common||>100 genes|
|Turn around times||8 weeks||16 weeks|
|Clinical sensitivity||All genes are associated with specific phenotype of panel||No specific phenotype needed; not all exons/genes covered|
|Gene coverage||Only genes included on panel are analyzed||Captures exomes indiscriminately|
Regardless of the debate, I think it’s safe to say that current NGS assays are not yet optimal for either: highly homologous regions (especially IonTorrent), triplet repeat expansions or structural variants. I’d love to hear how these technologies will operate here.