Although different instances of large GSVs commonly have approximately the same boundaries [10], with a correspondingly high probability of having the same or similar phenotypic effects, smaller GSV regions routinely feature a range of overlapping GSV calls of different sizes and locations [36], [37], thereby presenting a dilemma – in the absence of strong prior information to enable modelling of the effects of different GSV variants, how should a range of variants affecting a single locus be combined when testing for association with a phenotype of interest? One approach, analogous to methods such as the ‘cohort allelic sums test’ [38] developed for analysis of multiple rare sequence variants within a gene, is to treat a set of overlapping GSVs as functionally identical, effectively discounting the structural complexity, so that only a single hypothesis related to a putative functional effect is tested. Although perhaps appropriate where, for instance, all variants are predicted to have similar functional consequences due to directly disrupting or deleting a particular gene or due to affecting an intergenic region, this approach requires user intervention and is not
