Advantage of Using Allele-Specific Copy Numbers When Testing for Association in Regions with Common Copy Number Variants

Gaëlle Marenne, Stephen J. Chanock, Núria Malats, Emmanuelle Génin

Research output: Contribution to journalArticle

Abstract

Copy number variants (CNV) can be called from SNP-arrays; however, few studies have attempted to combine both CNV and SNP calls to test for association with complex diseases. Even when SNPs are located within CNVs, two separate association analyses are necessary, to compare the distribution of bi-allelic genotypes in cases and controls (referred to as SNP-only strategy) and the number of copies of a region (referred to as CNV-only strategy). However, when disease susceptibility is actually associated with allele specific copy-number states, the two strategies may not yield comparable results, raising a series of questions about the optimal analytical approach. We performed simulations of the performance of association testing under different scenarios that varied genotype frequencies and inheritance models. We show that the SNP-only strategy lacks power under most scenarios when the SNP is located within a CNV; frequently it is excluded from analysis as it does not pass quality control metrics either because of an increased rate of missing calls or a departure from fitness for Hardy-Weinberg proportion. The CNV-only strategy also lacks power because the association testing depends on the allele which copy number varies. The combined strategy performs well in most of the scenarios. Hence, we advocate the use of this combined strategy when testing for association with SNPs located within CNVs.

Original languageEnglish (US)
Article numbere75350
JournalPLoS One
Volume8
Issue number9
DOIs
StatePublished - Sep 10 2013
Externally publishedYes

Fingerprint

Single Nucleotide Polymorphism
Alleles
alleles
Testing
testing
Quality control
genotype
quality control
disease resistance
Genotype
inheritance (genetics)
Disease Susceptibility
Quality Control

ASJC Scopus subject areas

  • Agricultural and Biological Sciences(all)
  • Biochemistry, Genetics and Molecular Biology(all)
  • Medicine(all)

Cite this

Advantage of Using Allele-Specific Copy Numbers When Testing for Association in Regions with Common Copy Number Variants. / Marenne, Gaëlle; Chanock, Stephen J.; Malats, Núria; Génin, Emmanuelle.

In: PLoS One, Vol. 8, No. 9, e75350, 10.09.2013.

Research output: Contribution to journalArticle

Marenne, Gaëlle ; Chanock, Stephen J. ; Malats, Núria ; Génin, Emmanuelle. / Advantage of Using Allele-Specific Copy Numbers When Testing for Association in Regions with Common Copy Number Variants. In: PLoS One. 2013 ; Vol. 8, No. 9.
@article{a6f5f7f294614a0aa53af52e9c2e87fc,
title = "Advantage of Using Allele-Specific Copy Numbers When Testing for Association in Regions with Common Copy Number Variants",
abstract = "Copy number variants (CNV) can be called from SNP-arrays; however, few studies have attempted to combine both CNV and SNP calls to test for association with complex diseases. Even when SNPs are located within CNVs, two separate association analyses are necessary, to compare the distribution of bi-allelic genotypes in cases and controls (referred to as SNP-only strategy) and the number of copies of a region (referred to as CNV-only strategy). However, when disease susceptibility is actually associated with allele specific copy-number states, the two strategies may not yield comparable results, raising a series of questions about the optimal analytical approach. We performed simulations of the performance of association testing under different scenarios that varied genotype frequencies and inheritance models. We show that the SNP-only strategy lacks power under most scenarios when the SNP is located within a CNV; frequently it is excluded from analysis as it does not pass quality control metrics either because of an increased rate of missing calls or a departure from fitness for Hardy-Weinberg proportion. The CNV-only strategy also lacks power because the association testing depends on the allele which copy number varies. The combined strategy performs well in most of the scenarios. Hence, we advocate the use of this combined strategy when testing for association with SNPs located within CNVs.",
author = "Ga{\"e}lle Marenne and Chanock, {Stephen J.} and N{\'u}ria Malats and Emmanuelle G{\'e}nin",
year = "2013",
month = "9",
day = "10",
doi = "10.1371/journal.pone.0075350",
language = "English (US)",
volume = "8",
journal = "PLoS One",
issn = "1932-6203",
publisher = "Public Library of Science",
number = "9",

}

TY - JOUR

T1 - Advantage of Using Allele-Specific Copy Numbers When Testing for Association in Regions with Common Copy Number Variants

AU - Marenne, Gaëlle

AU - Chanock, Stephen J.

AU - Malats, Núria

AU - Génin, Emmanuelle

PY - 2013/9/10

Y1 - 2013/9/10

N2 - Copy number variants (CNV) can be called from SNP-arrays; however, few studies have attempted to combine both CNV and SNP calls to test for association with complex diseases. Even when SNPs are located within CNVs, two separate association analyses are necessary, to compare the distribution of bi-allelic genotypes in cases and controls (referred to as SNP-only strategy) and the number of copies of a region (referred to as CNV-only strategy). However, when disease susceptibility is actually associated with allele specific copy-number states, the two strategies may not yield comparable results, raising a series of questions about the optimal analytical approach. We performed simulations of the performance of association testing under different scenarios that varied genotype frequencies and inheritance models. We show that the SNP-only strategy lacks power under most scenarios when the SNP is located within a CNV; frequently it is excluded from analysis as it does not pass quality control metrics either because of an increased rate of missing calls or a departure from fitness for Hardy-Weinberg proportion. The CNV-only strategy also lacks power because the association testing depends on the allele which copy number varies. The combined strategy performs well in most of the scenarios. Hence, we advocate the use of this combined strategy when testing for association with SNPs located within CNVs.

AB - Copy number variants (CNV) can be called from SNP-arrays; however, few studies have attempted to combine both CNV and SNP calls to test for association with complex diseases. Even when SNPs are located within CNVs, two separate association analyses are necessary, to compare the distribution of bi-allelic genotypes in cases and controls (referred to as SNP-only strategy) and the number of copies of a region (referred to as CNV-only strategy). However, when disease susceptibility is actually associated with allele specific copy-number states, the two strategies may not yield comparable results, raising a series of questions about the optimal analytical approach. We performed simulations of the performance of association testing under different scenarios that varied genotype frequencies and inheritance models. We show that the SNP-only strategy lacks power under most scenarios when the SNP is located within a CNV; frequently it is excluded from analysis as it does not pass quality control metrics either because of an increased rate of missing calls or a departure from fitness for Hardy-Weinberg proportion. The CNV-only strategy also lacks power because the association testing depends on the allele which copy number varies. The combined strategy performs well in most of the scenarios. Hence, we advocate the use of this combined strategy when testing for association with SNPs located within CNVs.

UR - http://www.scopus.com/inward/record.url?scp=84883773762&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84883773762&partnerID=8YFLogxK

U2 - 10.1371/journal.pone.0075350

DO - 10.1371/journal.pone.0075350

M3 - Article

C2 - 24040408

AN - SCOPUS:84883773762

VL - 8

JO - PLoS One

JF - PLoS One

SN - 1932-6203

IS - 9

M1 - e75350

ER -