Electronic Theses and Dissertations

lei shi

Date

2020

Document Type

Dissertation

Degree Name

Doctor of Philosophy

Department

Mathematical Sciences

Committee Chair

Ebenezer George

Committee Member

Cheng Cheng

Committee Member

Dale Bowman

Committee Member

Stanley Pounds

Abstract

High throughput genome screening techniques are enabling researchers to interrogate human genome at single base pair level for their association with outcomes by genome wide association study (GWAS). However, it is usually challenging for GWAS to provide clear statistical conclusions at the gene level when multiple genomic features, either from same platform or different platforms, reside in the same gene. Traditionally a gene is considered as associated with an outcome when at least one genome feature within that gene is significantly associated with the outcome after adjusting for multiple genome-wide tests. Under that framework, only the most significant genome feature is used to determine the gene/outcome association. However adjustments for multiple testing impose a large penalty on single feature from high density arrays such as Affymetrix SNP6 arrays. Here we propose a procedure based on truncated and aggregated P values (TAP) to aggregate individual genome feature P-values within designated allele/gene. We then construct a hybrid permutation test to obtain a single P-value for the allele/gene in order to assess the overall association of the segment with clinical outcome.

Comments

Data is provided by the student.

Library Comment

Dissertation or thesis originally submitted to ProQuest

Download

Share

COinS