Limits...
Conditional genotype analysis: detecting secondary disease loci in linkage disequilibrium with a primary disease locus.

Thomson G, Valdes AM - BMC Proc (2007)

Bottom Line: In addition to the primary effect of the HLA DR locus, the effects of the other two HLA region simulated genes involved in disease were detected (gene C, 0 cM from DR, increases RA risk only in women; and gene D, 5.12 cM from DR, rare allele increases RA risk five-fold).No false negatives were found.Power calculations were performed.

View Article: PubMed Central - HTML - PubMed

Affiliation: Department of Integrative Biology, University of California, 3060 VLSB MC#3140, Berkeley, California 94720, USA. glenys@berkeley.edu

ABSTRACT
A number of autoimmune and other diseases have well established HLA associations; in many cases there is strong evidence for the direct involvement of the HLA class II peptide-presenting antigens, e.g., HLA DR-DQ for type 1 diabetes (T1D) and HLA-DR for rheumatoid arthritis (RA). The involvement of additional HLA region genes in the disease process is implicated in these diseases. We have developed a model-free approach to detect these additional disease genes using genotype data; the conditional genotype method (CGM) and overall conditional genotype method (OCGM) use all patient and control data and do not require haplotype estimation. Genotypes at marker genes in the HLA region are stratified and their expected values are determined in a way that removes the effects of linkage disequilibrium (LD) with the peptide-presenting HLA genes directly involved in the disease. A statistic has been developed under the hypothesis of no additional disease genes in the HLA region for the OCGM method and was applied to the Genetic Analysis Workshop 15 simulated data set of Problem 3, which mimics RA (answers were known). In addition to the primary effect of the HLA DR locus, the effects of the other two HLA region simulated genes involved in disease were detected (gene C, 0 cM from DR, increases RA risk only in women; and gene D, 5.12 cM from DR, rare allele increases RA risk five-fold). No false negatives were found. Power calculations were performed.

No MeSH data available.


Related in: MedlinePlus

Genetic association, linkage disequilibrium and application of conditional genotype analysis to chromosome 6 simulated data. A, Genotype associations of HLA region (chromosome 6) SNPs; B, linkage disequilibrium of SNPs 100 to 200 with the HLA DR4 allele; C, physical distribution of significant LD with chromosome 6 SNPs; D, application of the overall conditional genotype method (OCGM) to the HLA region data. Results refer to the average of 50 simulation replicates.
© Copyright Policy - open-access
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC2367484&req=5

Figure 1: Genetic association, linkage disequilibrium and application of conditional genotype analysis to chromosome 6 simulated data. A, Genotype associations of HLA region (chromosome 6) SNPs; B, linkage disequilibrium of SNPs 100 to 200 with the HLA DR4 allele; C, physical distribution of significant LD with chromosome 6 SNPs; D, application of the overall conditional genotype method (OCGM) to the HLA region data. Results refer to the average of 50 simulation replicates.

Mentions: On the simulated RA-like Problem 3 data we used one affected sib from all families (n = 1500) and all unrelated controls (n = 2000) and studied 674 SNPs on chromosome 6. First we tested for genotype associations of each SNP with disease, and found significant effects around the HLA region (Figure 1A); however the DR and C locus effects were intertwined. We then concentrated on SNPs 100–200 (the HLA region) in 50 replicates. A number of SNPs showed associations, and these correlated with estimated LD with the HLA DR4 allele simulated in the data (allele 3 of the simulation) (Figure 1B and 1C). Significant LD with DR4 extends from SNP 119 to 183, although except for these two extremes, the significant LD falls between SNPs 129 to 155.


Conditional genotype analysis: detecting secondary disease loci in linkage disequilibrium with a primary disease locus.

Thomson G, Valdes AM - BMC Proc (2007)

Genetic association, linkage disequilibrium and application of conditional genotype analysis to chromosome 6 simulated data. A, Genotype associations of HLA region (chromosome 6) SNPs; B, linkage disequilibrium of SNPs 100 to 200 with the HLA DR4 allele; C, physical distribution of significant LD with chromosome 6 SNPs; D, application of the overall conditional genotype method (OCGM) to the HLA region data. Results refer to the average of 50 simulation replicates.
© Copyright Policy - open-access
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC2367484&req=5

Figure 1: Genetic association, linkage disequilibrium and application of conditional genotype analysis to chromosome 6 simulated data. A, Genotype associations of HLA region (chromosome 6) SNPs; B, linkage disequilibrium of SNPs 100 to 200 with the HLA DR4 allele; C, physical distribution of significant LD with chromosome 6 SNPs; D, application of the overall conditional genotype method (OCGM) to the HLA region data. Results refer to the average of 50 simulation replicates.
Mentions: On the simulated RA-like Problem 3 data we used one affected sib from all families (n = 1500) and all unrelated controls (n = 2000) and studied 674 SNPs on chromosome 6. First we tested for genotype associations of each SNP with disease, and found significant effects around the HLA region (Figure 1A); however the DR and C locus effects were intertwined. We then concentrated on SNPs 100–200 (the HLA region) in 50 replicates. A number of SNPs showed associations, and these correlated with estimated LD with the HLA DR4 allele simulated in the data (allele 3 of the simulation) (Figure 1B and 1C). Significant LD with DR4 extends from SNP 119 to 183, although except for these two extremes, the significant LD falls between SNPs 129 to 155.

Bottom Line: In addition to the primary effect of the HLA DR locus, the effects of the other two HLA region simulated genes involved in disease were detected (gene C, 0 cM from DR, increases RA risk only in women; and gene D, 5.12 cM from DR, rare allele increases RA risk five-fold).No false negatives were found.Power calculations were performed.

View Article: PubMed Central - HTML - PubMed

Affiliation: Department of Integrative Biology, University of California, 3060 VLSB MC#3140, Berkeley, California 94720, USA. glenys@berkeley.edu

ABSTRACT
A number of autoimmune and other diseases have well established HLA associations; in many cases there is strong evidence for the direct involvement of the HLA class II peptide-presenting antigens, e.g., HLA DR-DQ for type 1 diabetes (T1D) and HLA-DR for rheumatoid arthritis (RA). The involvement of additional HLA region genes in the disease process is implicated in these diseases. We have developed a model-free approach to detect these additional disease genes using genotype data; the conditional genotype method (CGM) and overall conditional genotype method (OCGM) use all patient and control data and do not require haplotype estimation. Genotypes at marker genes in the HLA region are stratified and their expected values are determined in a way that removes the effects of linkage disequilibrium (LD) with the peptide-presenting HLA genes directly involved in the disease. A statistic has been developed under the hypothesis of no additional disease genes in the HLA region for the OCGM method and was applied to the Genetic Analysis Workshop 15 simulated data set of Problem 3, which mimics RA (answers were known). In addition to the primary effect of the HLA DR locus, the effects of the other two HLA region simulated genes involved in disease were detected (gene C, 0 cM from DR, increases RA risk only in women; and gene D, 5.12 cM from DR, rare allele increases RA risk five-fold). No false negatives were found. Power calculations were performed.

No MeSH data available.


Related in: MedlinePlus