Analysis of principal component based quantitative phenotypes for alcoholism

Principal component analysis was used to construct quantitative phenotypes for alcoholism. These were analyzed for linkage to genomic regions with a variance components approach. The four phenotypes considered were a factor describing medical symptoms of alcohol dependency, a factor describing a psychological profile correlated with susceptibility to alcoholism, monoamine oxidase B (MAOB) activity and an average measurement of the P3 component of event-related potentials (ERP) at the Fp electrode placements. One region (around marker GATA123C09 on chromosome 3) with suggestive evidence for linkage was detected for the P3 (Fp) measurement. For three of the four distinct phenotypes, modest evidence for linkage to a similar region (around marker ADH3 on chromosome 4) was found.

Incorporating larger families in identity-by-descent based linkage analysis

Genome scans for alcoholism susceptibility genes were carried out using identity-by-descent-based statistics for qualitative traits. We compared the results when 1) multipoint information was used for all families, where some had to be truncated, 2) multipoint information was used only for small families while large (untruncated) pedigrees were analyzed with a single-point approach, and 3) single-point analysis was used for all pedigrees. Differences between the methods were observed, but neither method could identify regions related to the susceptibility for alcoholism.

Systematic search for susceptibility genes in different populations

We analyzed the first replicate of each of the four simulated population samples from three distinct populations by linkage and association genome scans and could identify three regions with susceptibility loci for the disease: on chromosome 1, marker D1G024, with strong evidence for gene x environment interaction; or chromosome 3, around marker D3G045; and on chromosome 5, markers D5G035-D5G042. Our results were obtained without knowing the true disease model and are compared with this model in the discussion.

Genetic control of lipoprotein(a) concentrations is different in Africans and Caucasians

Lipoprotein(a) (Lp(a)) represents a quantitative trait in human plasma associated with atherothrombotic disease. Large variation in the distribution of Lp(a) concentrations exists across populations which is at present unexplained. Sib-pair linkage analysis has suggested that the apo(a) gene on chromosome 6q27 is the major determinant of Lp(a) levels in Caucasians. We have here dissected the genetic architecture of the Lp(a) trait in Africans (Khoi San, South African Blacks) and Caucasians (Austrians) by family/sib-pair analysis. Heritability estimates ranged from h2 = 51% in Blacks, h2 = 61% in Khoi San, to h2 = 71% in Caucasians. Analysis by a variance components model also demonstrated that the proportion of the total phenotypic variance explained by genetic factors is smaller in Africans (65%) than in Caucasians (74%). Importantly the sib-pair analysis clearly identified the apo(a) gene as the major locus in Caucasians which explained the total genetic variance. In the African samples the apo(a) gene accounted for only half the genetic variance. Together with previous results from population studies our data indicate that genetic control of Lp(a) levels seems to be distinctly different between Africans and Caucasians. In the former genetic factors distinct from the apo(a) locus and also non-genetic factors may play a major role.

Apolipoprotein E and Alzheimer disease: genotype-specific risks by age and sex

The distribution of apolipoprotein E (APOE) genotypes as a function of age and sex has been examined in a French population of 417 Alzheimer disease (AD) patients and 1,030 control subjects. When compared to the APOE epsilon3 allele, an increased risk associated with the APOE epsilon4 allele (odds ratio [OR] [epsilon4] = 2.7 with 95% confidence interval [CI] = 2.0-3.6; P < .001) and a protective effect of the APOE epsilon2 allele (OR[epsilon2] = 0.5 with 95% CI = 0.3-0.98; P = .012) were retrieved. An effect of the epsilon4 allele dosage on susceptibility was confirmed (OR[epsilon4/epsilon4] vs. the epsilon3/epsilon3 genotype = 11.2 [95% CI = 4.0-31.6]; OR[epsilon3/epsilon4] vs. the epsilon3/epsilon3 genotype = 2.2 [95% CI = 1.5-3.5]). The frequency of the epsilon4 allele was lower in male cases than in female cases, but, since a similar difference was found in controls, this does not lead to a difference in OR between sex. ORs for the epsilon4 allele versus the epsilon3 allele, OR(epsilon4), were not equal in all age classes: OR(epsilon4) in the extreme groups with onset at < 60 years or > 79 years were significantly lower than those from the age groups 60-79 years. In epsilon3/epsilon4 individuals, sex-specific lifetime risk estimates by age 85 years (i.e., sex-specific penetrances by age 85 years) were 0.14 (95% CI 0.04-0.30) for men and 0.17 (95% CI 0.09-0.28) for women.

Distribution of genome shared IBD by half-sibs: approximation by the Poisson clumping heuristic

The distribution of the proportion of genome shared identical by descent (IBD) by c half-sibs is calculated for a general c. With data availability the consideration of the chromosome as a whole rather than discrete loci becomes of increasing practical importance. Assuming the genomic continuum model, which allows for recombination, the crossover process in a chromosome pedigree is viewed as a continuous-time Markov random walk on the vertices of a hypercube with time parameter map distance along the chromosome. The desired distribution corresponds to the distribution of sojourn times of the process in a small set of vertices. The sojourn times are well approximated via the Poisson clumping heuristic. It is shown explicitly that there is no "equivalent" number of independently segregating loci that will yield the same results as with the genomic continuum model. Results are given for the human genome.

The probability distribution of the amount of an individual's genome surviving to the following generation

The probability that at least p% of an individual's genome is passed on collectively to his children is calculated. With data availability the consideration of the chromosome as a whole rather than discrete loci becomes of increasing practical importance. Assuming the genomic continuum model, which allows for recombination, the crossover process in a chromosome pedigree is viewed as a continuous-time Markov random walk on the vertices of a hypercube with time parameter map distance along the chromosome. The desired probability corresponds to the probability of sojourn times of the process in a small set of vertices, which are well approximated via the Poisson clumping heuristic. Results are given for the human genome. It is very likely that an individual with at least four children passes on at least 90% of his genome. There exists no "equivalent" number of independently segregating loci for this distribution.

Systematic search of susceptibility loci with methods using gametic disequilibrium

Susceptibility genes are identified for a simulated complex trait by a systematic genome search for linkage between disease and a genetic marker in the presence of gametic disequilibrium. The transmission/disequilibrium tests TDTa or TDTg for multiallelic markers compare transmitted and nontransmitted alleles or the genotypes formed by the two transmitted alleles and the genotypes formed by the two nontransmitted alleles, respectively. With these two tests we were able to identify the two markers D1G31 and D5G23. Under the simulating model these are in fact two susceptibility genes involved in the disease.

Comparing the power of linkage detection by the transmission disequilibrium test and the identity-by-descent test

The aim of this study is to compare the power of the transmission disequilibrium test (TDT) to that of the identity-by-descent (IBD) distribution test. The relative powers of these tests depend both on the underlying genetic model and on the available family data. Families with two affected sibs are always more informative than those with one affected child and one unaffected child. The IBD test is always more powerful in the first situation and, contrary to the TDT, is independent of the presence of gametic disequilibrium. When there is strong linkage disequilibrium, the TDT can be more powerful than the IBD test. In that case, linkage can be detected by the TDT even in families with only one affected child.

Statistical properties of the allelic and genotypic transmission/disequilibrium test for multiallelic markers

The transmission/disequilibrium test (TDT) is extended in two ways for a multiallelic marker: (1) to compare transmitted and nontransmitted alleles from a single heterozygous parent and (2) to compare genotypes formed by the two transmitted alleles and genotypes formed by the two nontransmitted alleles using the information on both parents, heterozygous or not, simultaneously.