Comprehensive Analysis of Subtelomeres by Genome Mapping and Sequencing

Detailed comprehensive knowledge of the structures of individual long-range telomere-terminal haplotypes are needed to understand their impact on telomere function, and to delineate the population structure and evolution of subtelomere regions. However, the abundance of large evolutionarily recent segmental duplications and high levels of large structural variations have complicated both the mapping and sequence characterization of human subtelomere regions. Here, I use a highly automated whole genome mapping technology in nano-channel arrays to analyze large terminal human chromosome segments extending from chromosome-specific subtelomere sequences through subtelomeric repeat regions to terminal (TTAGGG)n repeat tracts. I analyzed 154 human genomes from 26 populations to present a comprehensive look at human subtelomere structure and variation. The results catalog many novel long-range subtelomere haplotypes and determine the frequencies and contexts of specific subtelomeric duplicons on each chromosome arm, helping to clarify the currently ambiguous nature of many specific subtelomere structures as represented in the current reference sequence (hg38). The organization and content of some duplicons in subtelomeres appear to show both chromosome arm and population-specific trends. Based upon these trends, I estimate a timeline for the spread of these duplication blocks. The highly repetitive and variable short repeat elements (SREs) in subtelomeric regions present the significant challenges to complete the sequence assembly, as many additional copies of the blocks found through these methods are lacking in the current human reference. More often, the sequence reads (even the long PacBio reads) are aligned to the wrong locations in SRE regions as evidenced by high sequencing read coverage, and unusual allele distribution patterns in the aligned reads. Reads belonging to chromosomes lacking a reference copy of a block have been aligned to copies on other chromosomes that are present in the reference. I present a method for realigning and assigning these sequence reads to correct SREs. Long PacBio sequencing reads from each reference copy of these SRE blocks were pooled into consensus haplotypes based on allele distributions and combinations. Haplotypes were reassigned to the correct chromosome arms by evaluating their similarity to the reference copies of the block. I also used 10x Genomics Linked-Read sequencing data to assign the haplotypes to the extra SRE copies that are not present in hg38 by linking unassigned consensus haplotypes to the unique, single copy regions of the chromosome arms. This work also provides a better foundation for clinical diagnostics and understanding variants in subtelomeric regions.Ph.D., Biomedical Engineering -- Drexel University, 201