The Effect of Single Nucleotide Polymorphism Identification Strategies on Estimates of Linkage Disequilibrium

Abstract

At present there is tremendous interest in characterizing the magnitude and distribution of linkage disequilibrium (LD) throughout the human genome, which will provide the necessary foundation for genome-wide LD analyses and facilitate detailed evolutionary studies. To this end, a human high-density single-nucleotide polymorphism (SNP) marker map has been constructed. Many of the SNPs on this map, however, were identified by sampling a small number of chromosomes from a single population, and inferences drawn from studies using such SNPs may be influenced by ascertainment bias (AB). Through extensive simulations, we have found that AB is a potentially significant problem in estimating and comparing LD within and between populations. Specifically, the magnitude of AB is a function of the SNP discovery strategy, number of chromosomes used for SNP discovery, population genetic characteristics of the particular genomic region considered, amount of gene flow between populations, and demographic history of the populations. We demonstrate that a balanced SNP discovery strategy (where equal numbers of chromosomes are sampled from multiple subpopulations) is the optimal study design for generating broadly applicable SNP resources. Finally, we validate our theoretical predictions by comparing our results to publicly available data from ten genes sequenced in 24 African American and 23 European American individuals.

Impacts of Marker Class Bias Relative to Locus‐Specific Variability on Population Inferences in Chinook Salmon: A Comparison of Single‐Nucleotide Polymorphisms with Short Tandem Repeats and Allozymes

“…Because SNPs are often ascertained in batches, ascertainment bias probably acts in conjunction across several loci. The effects of ascertainment bias on inferences of some population parameters can be minimized by careful selection of the source and size of the ascertainment panel (Akey et al 2003), and this concept has been used to guide SNP discovery projects (e.g., Cappuccio et al 2006). The fact that we did not observe any impact of ascertainment bias on branching order, overall divergence, variance partitioning, or estimation of divergence among panel collections may be partially due to our relatively large and diverse panel.…”

Impacts of Marker Class Bias Relative to Locus‐Specific Variability on Population Inferences in Chinook Salmon: A Comparison of Single‐Nucleotide Polymorphisms with Short Tandem Repeats and Allozymes

“…Because SNPs are often ascertained in batches, ascertainment bias probably acts in conjunction across several loci. The effects of ascertainment bias on inferences of some population parameters can be minimized by careful selection of the source and size of the ascertainment panel (Akey et al 2003), and this concept has been used to guide SNP discovery projects (e.g., Cappuccio et al 2006). The fact that we did not observe any impact of ascertainment bias on branching order, overall divergence, variance partitioning, or estimation of divergence among panel collections may be partially due to our relatively large and diverse panel.…”

Association analysis of photoperiodic flowering time genes in west and central African sorghum [Sorghum bicolor (L.) Moench]

“…The mean r 2 of 0.18 was comparable to the study on sorghum [40] reporting r 2 > 0.1 but lower than the previous study on barley [41] which reported r 2 > 0.4. The variability in the range of r 2 estimates observed in our study can be due to the fact that linkage disequilibrium estimates vary according to the target genomic region as well as number of polymorphic sites [42,43]. Furthermore, because of limited coverage (small fragment size) of the CGs studied, it seems inappropriate to describe the decay of linkage disequilibrium along each CG.…”

Performance analysis of novel methods for detecting epistasis

“…In our experiments, missing data is simulated by removing 5% of genotype information randomly. Genotyping error is simulated using a directed-error model [54], so that, 5% of genotypes are selected and biased toward to one allele, unless it is already homozygous in the biased direction. Phenocopy is simulated such that 20% of cases are affected under a particular environmental condition, rather than genotype conditions.…”

Impacts of Marker Class Bias Relative to Locus‐Specific Variability on Population Inferences in Chinook Salmon: A Comparison of Single‐Nucleotide Polymorphisms with Short Tandem Repeats and Allozymes

“…Because SNPs are often ascertained in batches, ascertainment bias probably acts in conjunction across several loci. The effects of ascertainment bias on inferences of some population parameters can be minimized by careful selection of the source and size of the ascertainment panel (Akey et al 2003), and this concept has been used to guide SNP discovery projects (e.g., Cappuccio et al 2006). The fact that we did not observe any impact of ascertainment bias on branching order, overall divergence, variance partitioning, or estimation of divergence among panel collections may be partially due to our relatively large and diverse panel.…”

Association analysis of photoperiodic flowering time genes in west and central African sorghum [Sorghum bicolor (L.) Moench]

“…The mean r 2 of 0.18 was comparable to the study on sorghum [40] reporting r 2 > 0.1 but lower than the previous study on barley [41] which reported r 2 > 0.4. The variability in the range of r 2 estimates observed in our study can be due to the fact that linkage disequilibrium estimates vary according to the target genomic region as well as number of polymorphic sites [42,43]. Furthermore, because of limited coverage (small fragment size) of the CGs studied, it seems inappropriate to describe the decay of linkage disequilibrium along each CG.…”

Performance analysis of novel methods for detecting epistasis

“…In our experiments, missing data is simulated by removing 5% of genotype information randomly. Genotyping error is simulated using a directed-error model [54], so that, 5% of genotypes are selected and biased toward to one allele, unless it is already homozygous in the biased direction. Phenocopy is simulated such that 20% of cases are affected under a particular environmental condition, rather than genotype conditions.…”

Impacts of Marker Class Bias Relative to Locus‐Specific Variability on Population Inferences in Chinook Salmon: A Comparison of Single‐Nucleotide Polymorphisms with Short Tandem Repeats and Allozymes

“…Because SNPs are often ascertained in batches, ascertainment bias probably acts in conjunction across several loci. The effects of ascertainment bias on inferences of some population parameters can be minimized by careful selection of the source and size of the ascertainment panel (Akey et al 2003), and this concept has been used to guide SNP discovery projects (e.g., Cappuccio et al 2006). The fact that we did not observe any impact of ascertainment bias on branching order, overall divergence, variance partitioning, or estimation of divergence among panel collections may be partially due to our relatively large and diverse panel.…”

Association analysis of photoperiodic flowering time genes in west and central African sorghum [Sorghum bicolor (L.) Moench]

“…The mean r 2 of 0.18 was comparable to the study on sorghum [40] reporting r 2 > 0.1 but lower than the previous study on barley [41] which reported r 2 > 0.4. The variability in the range of r 2 estimates observed in our study can be due to the fact that linkage disequilibrium estimates vary according to the target genomic region as well as number of polymorphic sites [42,43]. Furthermore, because of limited coverage (small fragment size) of the CGs studied, it seems inappropriate to describe the decay of linkage disequilibrium along each CG.…”

Performance analysis of novel methods for detecting epistasis

“…In our experiments, missing data is simulated by removing 5% of genotype information randomly. Genotyping error is simulated using a directed-error model [54], so that, 5% of genotypes are selected and biased toward to one allele, unless it is already homozygous in the biased direction. Phenocopy is simulated such that 20% of cases are affected under a particular environmental condition, rather than genotype conditions.…”