Volume 6 Supplement 1
Identification of genes involved in alcohol consumption and cigarettes smoking
© de Andrade et al; licensee BioMed Central Ltd 2005
Published: 30 December 2005
We compared the results of quantitative linkage analysis using single-nucleotide polymorphisms and microsatellite markers and introduced a new screening test for multivariate quantitative linkage analysis using the Collaborative Study on the Genetics of Alcoholism data. We analyzed 115 extended non-Hispanic White families and tested for linkage using two phenotypes: the maximum number of drinks in a 24-hour period and the number of packs smoked per day for one year. Our results showed that the linkage signal increased using single-nucleotide polymorphisms compared with microsatellite markers and that the screening test gave similar results to that of the bivariate analysis, suggesting its potential use in reducing overall analysis time.
The Collaborative Study on the Genetics of Alcoholism (COGA) is a multicenter research program to detect and map susceptibility genes for alcohol dependence and related phenotypes. Numerous behavior measures were collected, two of which we considered for our study. The first is the maximum number of drinks in a 24 hour period (drink24), which can be considered a surrogate to alcoholism diagnosis and provides a quantitative measure to grade non-alcoholic individuals . The second measure is the number of packs smoked per day for one year (pakyrs). Since pakyrs is highly correlated with alcohol consumption , these two measures are good candidates for multivariate linkage analysis. The goals of our analysis were two-fold. First, we investigated the performance of a genome-wide scan using single-nucleotide polymorphisms (SNPs) relative to the microsatellite markers. Several studies have shown gains in information when SNPs are used for qualitative traits, but advantages and disadvantages of SNPs have not been explored with quantitative traits [3, 4]. Second, we evaluated a new screening test for multivariate quantitative linkage analysis using drink24 and pakyrs as two correlated behavioral measures. Previous linkage studies have investigated these measures individually [5, 6], but currently no study has considered them in a bivariate analysis. Bivariate quantitative linkage analyses have been shown to identify genes with small effects where these genes may be missed with univariate analyses. However, these multivariate linkage analyses are computationally intensive as the number of traits used in the analysis increases. The proposed screening test combines univariate linkage results to determine whether a bivariate linkage analysis might be beneficial.
The COGA data consisted of 143 extended families, a mixture of large and small families, with 1,350 family members with clinical and demographic data. Because these families consisted of different ethnicities, we analyzed the families that were white, non-Hispanic (WNH). A family was considered WNH if 75% of the reported ethnicity in the family was WNH; thus, our analyses were performed on 115 extended families. The phenotypes selected for the analysis were drink24 and pakyrs. Because drink24 and pakyrs measures have skewed distributions, a square root transformation (sqrt) was applied in both measures to normalize the distribution.
The microsatellite markers and the Illumina SNPs were each used for our analyses. For the Illumina SNPs, we removed SNPs that were in linkage disequilibrium (LD) with another SNP. We based our criteria for LD using r2, and the cut-off value of 0.4, which from our experience removed the effects of LD without a great loss of information. After dropping the SNPs in LD, a total of 350, 258, and 161 SNPs on chromosomes 1, 4, and 9, respectively, were used in our analyses. Multipoint identity-by-descent (MIBD) sharing among pairs of relatives was calculated for microsatellite and SNP markers using the SIMWALK2 software program 
Quantitative trait linkage analysis
For the quantitative linkage analysis, we used the locally developed SPLUS multic library. This is a new library based on the C++ multic program from ACT . For the analysis, we performed univariate and bivariate quantitative linkage analysis using a variance components (VC) approach. The details about univariate and multivariate quantitative linkage analysis are described in Amos  and de Andrade and Amos . Sqrt(pakyrs) and sqrt(drink24) were adjusted for age and sex in the linkage analyses.
To test for genetic linkage, a likelihood ratio test (LRT) was applied. Under the null hypothesis, the linked gene parameter(s) is (are) restricted to equal 0. The distributions of the univariate and bivariate linkage tests are a mixture of 1/2 χ02 and 1/2 χ12, and a mixture of 1/4 χ02, 1/2 χ12, and 1/4 χ32, respectively . In the univariate linkage analyses, we considered multipoint maximum LOD scores (MLS) ≥ 3.00 as statistically significant evidence of linkage, ≥ 2.00 as suggestive evidence, and ≥ 1.30 as tentative evidence of linkage . These MLS thresholds correspond to p-values of 0.0001, 0.001, and 0.007, respectively. To achieve levels of statistical significance in the bivariate linkage analysis comparable to the univariate thresholds, we calculated the threshold using a mixture of 1/4 χ02, 1/2 χ12, and 1/4 χ32. This calculation provided MLS ≥ 4.00 as statistically significant evidence of linkage (i.e., p ≤ 0.0001), ≥ 2.87 as suggestive evidence (i.e., p ≤ 0.001), and ≥ 2.06 as tentative evidence of linkage (p ≤ 0.007). We inferred evidence of chromosomal regions with pleiotropic effects when the bivariate MLS met the criteria for at least tentative evidence of linkage and its nominal p-value was less than the univariate maxima at the same location.
Let us assume k quantitative traits are represented by Y1, Y2, ..., Yk. For each trait a genome-wide scanning linkage analysis is performed using the VC quantitative trait approach. For each trait i, and genomic position j, the quantitative trait locus (QTL) variance component estimate (σ2ij) is estimated with its standard error. Our hypothesis for the proposed screening test is: if there is a gene with pleiotropic effects, its QTL VC should be incremented in an additive manner using combinations of correlated traits by simply adding its respective univariate QTL VC. Let σ2ijk be the QTL VC for trait i, position j on chromosome k. The null hypothesis is that there is no pleiotropic effect at position j on chromosome k, i.e., H0: = 0 ∀ i, j, k, which is equivalent to H0: . The alternative hypothesis is H1: σ2 ijk > 0 for any i, j, k. The test statistic will be , where is the maximum likelihood estimator (MLE) of σ2 ijk . Under H0, E(σ2 ijk ) = 0, ∀ i, j, k, S ijk ~ 1/2 N (0, 1). By assuming the S ijk values are independent, , where T is the number of traits. Consequently by squaring and standardizing T jk , .
In our analyses using microsatellite markers, tentative and suggestive evidence of linkage were found on chromosomes 1, 2, 8, and 14 for sqrt(pakyrs) and on chromosomes 10 and 13 for sqrt(drink24). Bergen et al. identified several regions for sqrt(pakyrs) in the COGA sample, among chromosomes 2 (~10 cM) and 14 (~68 cM) . Straub et al. identified several linkage regions for nicotine dependence in a sample from Christchurch, New Zealand within chromosome 2 (~150 cM, LOD = 1.5) . Saccone et al. identified a susceptibility locus on chromosome 4 (~120 cM, LOD = 3.5) for drink24 . In our analysis using SNP markers, we observed an increase in the LOD scores and suggestive evidence of linkage on chromosomes 1 and 4 for sqrt(drink24) that was not observed using microsatellite markers. No evidence of a pleiotropic effect was found between sqrt(pakyrs) and sqrt(drink24). Our screening test is a computationally time-saving approach that can be used to determine which regions should be analyzed using a multivariate approach. However, significant results of the screening test may be misleading because the results may be driven by only one trait rather than several traits. Thus, careful evaluation of the univariate linkage results and the screening test is necessary.
During our analyses several difficulties arose when SNPs were used in quantitative trait linkage analysis. First, the only software that could specifically handle pedigrees of large size was SIMWALK2 ; however, it was computationally intensive to estimate the MIBDs. Second, in order to calculate the MIBD for 350 SNPS on chromosome 1, we had to break the 350 SNPs into 10 groups of 35 SNPs and then combine the results of the linkage analyses.
We observed evidence of linkage on chromosome 4 for alcohol consumption using SNPs; this linked region was in the same region previously identified by Saccone et al. . Furthermore, using SNPs, we also observed several suggestive regions for linkage to sqrt(pakyrs) and sqrt(drink24) not previously identified. The proposed screening test for multivariate quantitative trait linkage analysis also showed its potential application in this data. Our experience using large extended families and many SNPs suggest that software limitations are an issue when contemplating genome-wide linkage scans.
Collaborative Study on the Genetics of Alcoholism
Likelihood ratio test
Maximum likelihood estimator
Maximum LOD score
Quantitative trait locus
This research was partially funded by NIH grants R01HL71917 and CA94919.
- Bierut LJ, Saccone NL, Rice JP, Goate A, Foroud T, Edenberg H, Almasy L, Conneally PM, Crowe R, Hesselbrock V, Li TK, Nurnberger J, Porjesz B, Schuckit MA, Tischfield J, Begleiter H, Reich T: Defining alcohol-related phenotypes in humans. Alcohol Res Health. 2002, 26: 208-213.PubMedGoogle Scholar
- Schiffman S, Balabanis BA: Associations between alcohol and tobacco. Alcohol and Tobacco : From Basic Science to Clinical Practice. Edited by: Fertig JB, Allen JP. 1995, Bethesda, MD: National Institutes of Health, 17-36.Google Scholar
- Middleton FA, Pato MT, Gentile KL, Morley CP, Zhao X, Eisener AF, Brown A, Petryshen TL, Kirby AN, Medeiros H, Carvalho C, Macedo A, Dourado A, Coelho I, Valente J, Soares MJ, Ferreira CP, Lei M, Azevedo MH, Kennedy JL, Daly MJ, Sklar P, Pato CN: Genomewide linkage analysis of bipolar disorder by use of a high-density single-nucleotide-polymorphism (SNP) genotyping assay: a comparison with microsatellite marker assays and finding of significant linkage to chromosome 6q22. Am J Hum Genet. 2004, 74: 886-897. 10.1086/420775.PubMed CentralView ArticlePubMedGoogle Scholar
- John S, Shephard N, Liu G, Zeggini E, Cao M, Chen W, Vasavda N, Mills T, Barton A, Hinks A, Eyre S, Jones KW, Ollier W, Silman A, Gibson N, Worthington J, Kennedy GC: Whole-genome scan, in a complex disease, using 11,245 single-nucleotide polymorphisms: comparison with microsatellites. Am J Hum Genet. 2004, 75: 54-64. 10.1086/422195.PubMed CentralView ArticlePubMedGoogle Scholar
- Saccone NL, Kwon JM, Corbett J, Goate A, Rochberg N, Edenberg HJ, Foroud T, Li TK, Begleiter H, Reich T, Rice JP: A genome screen of maximum number of drinks as an alcoholism phenotype. Am J Med Genet. 2000, 96: 632-637. 10.1002/1096-8628(20001009)96:5<632::AID-AJMG8>3.0.CO;2-#.View ArticlePubMedGoogle Scholar
- Bergen AW, Yang XR, Bai Y, Beerman MB, Goldstein AM, Goldin LR, Framingham Heart Study: Genomic regions linked to alcohol consumption in the Framingham Heart Study. BMC Genetics. 2003, 4 (Suppl 1): S101-10.1186/1471-2156-4-S1-S101.PubMed CentralView ArticlePubMedGoogle Scholar
- Sobel E, H Sengul H, Weeks DE: Multipoint estimation of identity-by-descent probabilities at arbitrary positions among marker loci on general pedigrees. Hum Hered. 2001, 52: 121-131. 10.1159/000053366.View ArticlePubMedGoogle Scholar
- de Andrade M, Krushkal J, Yu L, Zhu D, Amos CI: ACT – A computer package for analysis of complex traits [abstract]. Am J Hum Genet. 1998, 63: A297-10.1086/301991.View ArticleGoogle Scholar
- Amos CI: Robust variance-components approaches for assessing genetic linkage in pedigrees. Am J Hum Genet. 1994, 54: 535-543.PubMed CentralPubMedGoogle Scholar
- de Andrade M, Amos CI: Multivariate linkage analysis. Biostatistical Genetics and Genetic Epidemiology. Edited by: Elson R, Olson J, Palmer L. 2002, New York: John Wiley & Sons, 564-568.Google Scholar
- Self SG, Liang K-L: Asymptotic properties of maximum likelihood estimators and likelihood ratio tests under nonstandard conditions. J Am Stat Assoc. 1987, 82: 605-610. 10.2307/2289471.View ArticleGoogle Scholar
- Morton NE: Significance levels in complex inheritance. Am J Hum Genet. 1998, 62: 690-697. 10.1086/301741.PubMed CentralView ArticlePubMedGoogle Scholar
- Bergen AW, Korczak JK, Weissbecker KA, Goldstein AM: A genome wide-search for loci contributing to smoking and alcoholism. Genet Epidemiol. 1999, 17 (Suppl l): S55-S60.View ArticlePubMedGoogle Scholar
- Straub RE, Sullivan PF, Ma Y, Myakishev MV, Harris-Kerr C, Wormley B, Kadambi B, Sadek H, Silverman MA, Webb BT, Neale MC, Bulik CM, Joyce PR, Kendler KS: Susceptibility genes for nicotine dependence: a genome scan and followup in an independent sample suggest that regions on chromosomes 2, 4, 10, 16, 17 and 18 merit further study. Mol Psychiatry. 1999, 4: 129-144. 10.1038/sj.mp.4000518.View ArticlePubMedGoogle Scholar
This article is published under license to BioMed Central Ltd. This is an open access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.