Structural variants, much ado about nothing?

Structural variants (SVs) were one of the first types of genetic variants to be studied because of their large size, which allowed them to be identified by cytogenetic methods. However, in humans they were traditionally considered to be either benign rearrangements with no known consequences or the cause of genomic disorders of large phenotypic effect [1–3]. Thanks to the advances in sequencing and genotyping technologies, the interest quickly shifted to single nucleotide polymorphisms (SNPs), with a golden era of genome-wide association studies (GWAS). Approximately 10 years ago, the discovery that structural variation was much more frequent than previously thought [4–6], opened a new trend in human genetics and a decade of intense research. This resulted in the identification of thousands of SVs, the determination of the frequency and distribution of many of them across human populations and some insight into their functional consequences and their evolutionary history (see for example [7]). Although the precise amount of structural variation is uncertain, the consensus is that it spans a higher proportion of the human genome than SNPs [8]. Moreover, SVs could also have a wider range of effects, going from simple gene mutations, to dosage alteration of functional elements, predisposition to other rearrangements or inhibition of recombination, like in the case of inversions. Therefore, since the beginning, there were high expectations that SVs might explain at least part of the missing heritability in humans, and a race was set to find their contribution to common diseases and complex traits [9–11]. Unfortunately, only a few associations between SVs and diseases have been reported, and there is a sense of disappointment about the functional impact of these changes [12]. However, the problem is that accurate detection of SVs is difficult, and our understanding of structural variation is still limited. In particular, current SV catalogs include a significant number of false positives, and many changes in complex regions with large segmental duplications are probably missing, which complicates making reliable conclusions. Thus, the field is ready for more precise and detailed studies, in which there is a careful and systematic characterization of these variants [13–15]. This special issue of ‘Briefings in Functional Genomics’ provides a global overview of the functional impact in humans of different types of SVs, from the most studied ones, copy number variants (CNVs), to others not so well known, such as inversions. The threshold of what is considered as structural variation is somehow arbitrary, and its size has been driven down by the increase in resolution of the techniques. The reviews included here take the most accepted point of view and generally focus on variants ranging from 50–100 bases to megabases. In addition, we have tried to summarize a diverse range of evidence on the functional effects of SVs, including disease associations, effects on genes and gene expression and their evolutionary role. The first article by Escaramís et al. [16] introduces an historical perspective of how SVs are defined, how they are generated and the methods used for their detection. Then, the next three reviews focus on CNVs and other types of SVs and human diseases. Iyer and Girirajan [17] provide a summary of rare CNVs involved in neurodevelopmental disorders that include some of the large deletions and duplications behind the first recurrent genomic disorders, their genetic architecture and the different strategies to study the casual genes, using the information both of human patients and animal models. Usher and McCarroll [18] write about complex multiallelic CNVs, which have not been well assessed before and have a great potential to affect phenotypes and disease susceptibility, and how it is possible to overcome the limitations of their study. In the last of these reviews, Tubio [19] makes an exhaustive account of the mutational processes underlying somatic SVs associated with different types of cancer. Another important evidence of the functional impact of a genetic variant is its effect on gene expression. Hence, Gamazon and Stranger [20] review the mechanisms by which CNVs can affect gene-expression levels and current findings on the regulatory roles of CNVs. Evolutionary analysis could also be a powerful alternative to illustrate the functional importance of genetic changes, and Radke and Lee [21] discuss about the adaptive potential of SVs in humans and other mammals, providing cases of different SVs affecting selected traits. Finally, Puig et al. [22] describe the current knowledge about human genome inversions and their impact at different levels, from individual mutations, association with diseases, gene expression and evolution. It is important to note that several of the types of variants described here have been poorly or unreliably analyzed before. For example, somatic mutations need to be checked in a tissue-by-tissue basis and not only by analyzing blood-related cells. In a similar way, multiallelic CNVs and inversions could appear recurrently, which means that they have a complex relationship with SNPs, and their effects have been largely overlooked by GWAS screens so far. Therefore, this is likely just the tip of the iceberg, and there might be much to discover yet about the functional impact of SVs in the human genome.