TY - JOUR TI - A statistical method for genotypic association that is robust to sequencing misclassification DO - https://doi.org/doi:10.7282/T3XS5Z9X PY - 2017 AB - In analyzing human genetic disorders, association analysis is one of the most commonly used approaches. However, there are challenges with association analysis, including differential misclassification in data that inflates the false-positive rate. In this thesis, I present a new statistical method for testing the association between disease phenotypes and multiple single nucleotide polymorphisms (SNPs). This method uses next-generation sequencing (NGS) raw data and is robust to sequencing differential misclassification. By incorporating expectation-maximization (EM) algorithm, this method computes the test statistic and estimates important parameters of the model, including misclassification. By performing simulation studies, I report that this method maintains correct type I error rates and may obtain high statistical power. KW - Microbiology and Molecular Genetics KW - Genomics--Data processing KW - Computational biology LA - eng ER -