Computational Assessment of Genetic Variation beyond Single Nucleotide Changes

Advances in sequencing technology have greatly reduced the costs incurred in collecting raw sequencing data and researchers now have access to very large datasets of genomic alterations. Computational tools are necessary in order to interpret and discover biologically relevant genetic variation from sequencing data. Current computational tools, however, have overwhelmingly focused on single nucleotide changes. Much less work has been devoted to computational tools to prioritize insertion and deletion variants and chromosomal abnormalities. Insertion/deletion variants (indels) alter protein sequence and length, yet are highly prevalent in healthy populations, presenting a pressing need for bioinformatics classifiers. Chromosomal abnormalities can produce a wide range of genetic disorders including in miscarriages, developmental disorders, and carcinogenesis. While numerous tools have been developed to detect chromosomal abnormalities, these tools have limited utility at lower cell admixtures. In this dissertation, I focus on the development of computational approaches beyond single nucleotide variants. I introduce a novel computational approach to assess indels variants (Chapters 2-3). I compare this method to existing computational approaches and investigate potential ways to improve indel prediction. Next, I develop a bioinformatics approach entitled WALDO (Within-sample AneupLoidy DiscOvery) specifically designed to detect chromosomal abnormalities as well as microsatellite instability (Chapters 4-6)