Grey zone. Structural genomic variants or copy number variations (CNV) can be reliably assessed using array comparative genomic hybridization (array CGH) or Single Nucleotide Polymorphism (SNP) arrays. However, for deletions or duplications smaller than 50-100 kB, these technologies have a poor detection rate with many false positive and false negative findings unless platforms are used that target specific candidate regions. Exome analysis, on the other hand, is capable of assessing genetic variation reliably on the single base-pair level. Between both technologies, there is a grey zone of structural genomic variants that are difficult to detect; CNVs smaller than 50 kB are often difficult to assess, and the extent and pathogenic role of these small CNVs is unclear. Now, a recent paper in the American Journal of Human Genetics manages to detect small CNVs through exome data. Their analysis in patients with autism, parents, and unaffected siblings suggests a contribution of small inherited CNVs to the overall autism risk.
Structural genomic variants in autism. While the role of structural genomic variants in various neurodevelopmental disorders is well established in 2013, it took quite some effort to get there. CNVs were first systematically assessed in large autism cohorts before the same technology was used in other neurodevelopmental disorders. The finding that large de novo copy number variants represented a significant risk factor for autism paved the way for a successful journey of gene discovery for neurodevelopmental disorders, which culminated in the recent exome studies in autism, intellectual disability, schizophrenia, and epilepsy. Amongst the various genetic alterations identified in these disorders, large structural genomic variations proved to be the most frequently encountered genetic changes. In patients with autism, for example, up to 10% carry a pathogenic de novo structural genomic variation. However, there is a limit regarding the resolution of available platforms.
Going smaller – CNVs through exomes. While large structural genomic variants can easily be assessed through various platforms, the detection of variants smaller than 50-100 kB is problematic. For both array CGH and SNP arrays, variants of this size will need to be called with a small number of probes, introducing statistical noise, which leads to a poor rate of true positive findings. For SNP arrays, the resolution with regards to small CNVs may be particularly poor, especially when DNA of suboptimal quality is used. Therefore, the grey zone of small CNVs is virtually unexplored. Neither the abundance of small CNVs nor their role in human disease is understood. Accordingly, there are several attempts to chart this unexplored territory. One promising way of exploring small CNVs is CNV analysis through exome data. In principle, this technology aims to use read depth as a way to identify duplications and deletions. Read depth refers to the number of reads that cover a given base pair. If a given DNA segment is deleted or duplicated, the number of reads will be different. This method has been refined over the recent years. It is now ready to be applied to research questions.
Small CNVs in autism. Krumm and collaborators analyzed available genetic data from the Simons Simplex Collection (SSC), a resource for autism genetics. The families included in the current study include patients with autism, their unaffected parents, and one unaffected sibling. The availability of sequence information of the unaffected sibling allowed the authors to compare the over transmission of small structural genomic variants to the affected sibling compared to the unaffected sibling. However, prior to establishing the risk conferred by small CNVs, some basic footwork needed to be performed to establish the validity of their CNV calling method. Krumm and collaborators used an algorithm called CoNIFER, which extracts CNV information through exome read depth. Despite the fact that this method might be seen as imprecise or unreliable, it worked amazingly well and discovered roughly twice as many small CNVs as a commercially available array CGH platform. The identified CNVs had an average size of 19 kB, and only 7% of CNVs detected through the calling algorithm turned out to be false positive findings. This is much, much better than you would naively expect based on other early experiences with CNV calling algorithms. Given this plethora of newly identified small deletions and duplications, what risk do they contribute to autism?
The risk of small CNVs. Krumm and collaborators analyzed 411 families from the SSC and found a higher frequency of small CNVs in probands compared to unaffected siblings, even though the effect was relatively minor. When expressed as the CNV “burden”, the increase in affected siblings was 1.2, and the burden for gene-containing small CNVs was only slightly higher. However, these findings were significant. Additional analysis by the authors suggested that the risk conferred by these small CNVs was independent of the risk conferred by large CNVs and de novo mutations. Accordingly, small inherited CNVs represent a novel, independent cause of genetic morbidity for neurogenetic disorders.
The role of small CNVs in epilepsy. Given the availability of exome data for several hundreds of trios with epileptic encephalopathies, a similar analysis may be performed to assess the role of this genetic mechanism in seizure disorders. Identifying recurrent small CNVs may help narrow down the margin of unexplained cases and help identify novel candidate genes. Particularly small CNVs may only contain single genes, which may help identify novel genetic mechanisms for epilepsy.