Comparison of single-nucleotide polymorphisms and microsatellite markers for linkage analysis in the COGA and simulated data sets for Genetic Analysis Workshop 14: Presentation Groups 1, 2, and 3

Abstract

The papers in presentation groups 1-3 of Genetic Analysis Workshop 14 (GAW14) compared microsatellite (MS) markers and single-nucleotide polymorphism (SNP) markers for a variety of factors, using multiple methods in both data sets provided to GAW participants. Group 1 focused on data provided from the Collaborative Study on the Genetics of Alcoholism (COGA). Group 2 focused on data simulated for the workshop. Group 3 contained analyses of both data sets. Issues examined included: information content, signal strength, localization of the signal, use of haplotype blocks, population structure, power, type I error, control of type I error, the effect of linkage disequilibrium, and computational challenges. There were several broad resulting observations. 1) Information content was higher for dense SNP marker panels than for MS panels, and dense SNP markers sets appeared to provide slightly higher linkage scores and slightly higher power to detect linkage than MS markers. 2) Dense SNP panels also gave higher type I errors, suggesting that increased test thresholds may be needed to maintain the correct error rate. 3) Dense SNP panels provided better trait localization, but only in the COGA data, in which the MS markers were relatively loosely spaced. 4) The strength of linkage signals did not vary with the density of SNP panels, once the marker density was approximately 1 SNP/cM. 5) Analyses with SNPs were computationally challenging, and identified areas where improvements in analysis tools will be necessary to make analysis practical for widespread use.

Genome-wide linkage scan for contraction velocity characteristics of knee musculature in the Leuven Genes for Muscular Strength Study

“…Compared with our previous analyses (18) with microsatellites markers (average marker distance 4.81 cM for chromosome 12), the current study used the Illumina predesigned SNPbased Linkage IVb panel with an average marker distance of 0.56 cM for chromosome 12. Although this SNP panel induces higher genetic informativeness than microsatellite markers (13,47), differences between present (n ϭ 283, Illumina SNP panel) and previous findings (n ϭ 367, microsatellite markers; Applied Biosystems) (18) are probably related to sample size differences or region-specific informativeness of closely chosen microsatellite markers vs. SNP markers (mean heterozygosity of SNP markers was 44.2% on chr12). Of note, also several other overlapping/neighboring linkage regions could be detected between the present study and our genome-wide scan for maximum and length-dependent knee muscle strength (12): region 2p25.3 (TVRE) with 2p24.2 for isometric torque at 30°fl exion; region 10q25.3 (TVSE) with 10q26.13 for torquelength flexion; region 2q14.3 (TVSE) with 2q14.2 for musclebone area of the midthigh; region 12q24.32 (TVSE) with 12q24.32 and 12q24.31 for torque-length flexion and extension, respectively; and 15q23 (TVSE) with 15q23 for isometric torque at 30°flexion.…”

Genome-wide Linkage Analysis with Clustered SNP Markers

“…2 In these comparisons, the technical performance measures such as call rates and genotyping error rates of the different SNP genotyping methods seem to be comparable. However, as pointed out by Wilcox et al, 2 1 of the main challenges in parametric linkage analysis in the future is the computational handling of the large amount of data produced from the analysis of large pedigrees. This was also the case in our project, and here we report our experiences with the newly launched applied Biosystems LMS 4K.…”

Association mapping: methodologies, strategies, and issues

“…With SNP markers so densely spaced, this assumption will be violated in many SNP-based linkage studies. (This issue is discussed further by Wilcox et al [2005].) Contributors' assessments of the effects of LD on their SNP-based linkage scans yielded mixed results.…”