Structural variation and Alu elements in diverse human populations

Abstract

The study of human variation has long been of interest in the effort to understand changes in the human genome that have implications for health and disease. The advent of high-throughput sequencing has allowed for the large-scale investigation of genomic variation across the world, revealing that much of this is population-specific. However, such studies have focused on a limited number of populations, most of which have surveyed primarily European individuals. Should patterns of genomic variation not be described within underrepresented populations with adequate diverse representation, these communities will be left behind as genomic medicine progresses and becomes more personalised. Currently, this includes Indigenous Australians, whose genomic diversity is not well characterised.

Structural variation (SV), large rearrangements within genomes, significantly contributes to the human genomic variation. Compared to short variants, however, their mutational processes, population frequencies and functional implications are still unknown, particularly in ancestrally diverse and minority populations. Alu elements, which have been implicated in contributing to structural variation, are similarly understudied.

This thesis first characterises germline structural variation in 462 Indigenous Australians from four communities. This is described in the context of 270 individuals from 128 ancestrally diverse population groups represented in the Simons Genome Diversity Project and 24 individuals representing 24 populations in the 1000 Genomes Project. Here, we document 102,925 SVs, primarily composed of deletions, duplications, and inversions. We report an enrichment in the number of duplications within Indigenous Australian individuals, and that Indigenous Australians and Africans harbour most of the novel structural variation that has not been previously reported. This highlights the need to survey such diverse populations more widely and deeply. Finally, we report that variants that are globally rare and locally common are specific within a subpopulation/community, suggesting that human variation needs to be surveyed at a community level to improve population reference data in Indigenous Australians.

Using a novel de novo assembly-based approach for short-read whole-genome sequence data, an in-depth characterisation of Alu polymorphisms within these individuals is also discussed. We first describe the sensitive and specific nature of our approach, finding that 99.4% of our callset overlaps with reference Alu elements, and that 93-96% of reference Alus are detected. Using this, we describe population Alu polymorphisms, including with the relatively understudied AluS class and Alu deletions, and highlight the need to survey such elements to comprehensively understand Alu contributions to human genomic diversity. Using both our SV and Alu callsets, we also describe the role of Alu in contributing to a significant portion (43%) of structural variants in the human genome, highlighting the need to understand the role of Alu in SV formation. Finally, we reveal novel insights into the role of Alu in the formation of inversions through an in-depth description of mechanistic processes underlying Alu-mediated retrotransposon recombination-mediated inversions (RRMI).

This thesis completes a picture of genomic variation within humans through its population-scale examination of Australian individuals, revealing population-specific patterns, functional implications, and their relationship with diverse populations. It catalogues genomic information critical for the generation of population reference panels and provides novel insights into structural variant and Alu element mechanisms that are relevant for health and disease. Show more