Visualizing shifts in Founder Assignment for an Admixed Population of Mice caused by a New Genomic Build

Populations of inbred laboratory mice are highly desirable as model organisms due to their easily reproducible (isogenic) genomes. A combination of SNP genotyping arrays and a hidden Markov model (HMM) algorithm have been used successfully to accurately determine the genetic architecture of inbred mice. This HMM algorithm, known as the Founder Assignment algorithm, performs statistical modeling on genotyping data to calculate the probability that a given segment of a mouse’s genome was inherited from a particular founding strain of the population. However, the predictions generated by this algorithm are sensitive to changes in the genomic positions of the SNP markers used for genotyping. Although many groups have worked to improve the accuracy and resolution of HMM algorithms, few have considered how new assemblies of the genome, in which entire sets of markers may be reordered or removed, affect HMM predictions. In this study, the Founder Assignment algorithm was adapted to work with a new assembly (Build 38) of the mouse genome. The algorithm was then run on a collection of genotyping data from approximately 400 mice. A new visualization tool, which more clearly displays the HMM probabilities, was developed in order to identify regions where the algorithm assigned different founders between Build 38 and the previous build. In the future, these insights will help us design versions of the algorithm to work with newer versions of the genotyping array and with high-throughput sequencing data.Bachelor of Scienc