|
Activity Number:
|
178
|
|
Type:
|
Topic Contributed
|
|
Date/Time:
|
Monday, August 4, 2008 : 2:00 PM to 3:50 PM
|
|
Sponsor:
|
Section on Bayesian Statistical Science
|
| Abstract - #301324 |
|
Title:
|
Detection of Copy Number Variations from High-Density SNP Arrays: An Integrated Bayesian Hidden Markov Model Approach Incorporating Pedigree Information
|
|
Author(s):
|
Zhen Chen*+ and Mahlet Tadesse and Kai Wang and Mingyao Li
|
|
Companies:
|
University of Pennsylvania and Georgetown University and University of Pennsylvania and University of Pennsylvania
|
|
Address:
|
, Philadelphia, PA, ,
|
|
Keywords:
|
Hidden markov model ; MCMC ; Copy number variations ; Bayesian ; Genome ; Model uncertainty
|
|
Abstract:
|
Copy number variations (CNVs) refer to gains and losses of genomic elements compared to a reference genome assembly. Studies have demonstrated the heritability of CNVs but few incorporated family structures. We develop an integrated Bayesian approach that aims to incorporate family relationships when inferring CNVs. We assume copy number sequence along the chromosome follows a Markov model with transition probabilities dependent on genetic distances between adjacent SNPs. We also allow for de novo events in offspring's CNV calls and use another HMM to account for the dependence with neighboring SNPs in the same de novo CNV region. Our approach yields posterior distributions of CNV configurations, thus providing an uncertainty measure for the inferred CNVs. We evaluate the performance of the method by applying it to simulated datasets and the CEU trio data from HapMap.
|
