To submit an abstract for the meeting, please complete all of the following fields.

Type full text of the abstract in the box below:

If you need to include special characters, such as Greek letters, use formatting tags for them as shown in the table of formatting and special characters. characters left
Keywords: (choose up to three) Affected relative pairs Alzheimer Disease Ascertainment Association Autoimmune disorders Bayesian analysis Bioinformatics Cancer Cardiovascular risk factors and hypertension Case-only design Causation Data mining Diabetes Familial aggregation and segregation analysis Fine mapping Gene expression arrays Gene expression patterns Gene-environment interaction Genome scan Genomic variation Haplotype analysis Heritability Heterogeneity, Homogeneity Inbreeding, isolate populations Liability Linkage analysis Linkage and association Markov Chain Monte Carlo methods Maximum likelihood methods Missing data, genotyping errors Multilocus analysis Multiple marker disequilibrium analysis Multivariate phenotypes Mutation analysis Pedigree likelihood Pharmacogenetics Population stratification Quantitative trait analysis Sample size Sequential study design Single nucleotide polymorphism Transmission / disequilibrium test Variance components No Keyword Affected relative pairs Alzheimer Disease Ascertainment Association Autoimmune disorders Bayesian analysis Bioinformatics Cancer Cardiovascular risk factors and hypertension Case-only design Causation Data mining Diabetes Familial aggregation and segregation analysis Fine mapping Gene expression arrays Gene expression patterns Gene-environment interaction Genome scan Genomic variation Haplotype analysis Heritability Heterogeneity, Homogeneity Inbreeding, isolate populations Liability Linkage analysis Linkage and association Markov Chain Monte Carlo methods Maximum likelihood methods Missing data, genotyping errors Multilocus analysis Multiple marker disequilibrium analysis Multivariate phenotypes Mutation analysis Pedigree likelihood Pharmacogenetics Population stratification Quantitative trait analysis Sample size Sequential study design Single nucleotide polymorphism Transmission / disequilibrium test Variance components No Keyword Affected relative pairs Alzheimer Disease Ascertainment Association Autoimmune disorders Bayesian analysis Bioinformatics Cancer Cardiovascular risk factors and hypertension Case-only design Causation Data mining Diabetes Familial aggregation and segregation analysis Fine mapping Gene expression arrays Gene expression patterns Gene-environment interaction Genome scan Genomic variation Haplotype analysis Heritability Heterogeneity, Homogeneity Inbreeding, isolate populations Liability Linkage analysis Linkage and association Markov Chain Monte Carlo methods Maximum likelihood methods Missing data, genotyping errors Multilocus analysis Multiple marker disequilibrium analysis Multivariate phenotypes Mutation analysis Pedigree likelihood Pharmacogenetics Population stratification Quantitative trait analysis Sample size Sequential study design Single nucleotide polymorphism Transmission / disequilibrium test Variance components
Here is an example abstract: On Maximizing Multipoint Heterogeneity Lod Statistics over the Genetic Parameter Space Constrained by Sample Prevalence C.-H. Chen(1), S.K. Nath(1), S.J. Finch(2), N.R. Mendell(2), D. Gordon(3), G.C. Wedig(1), R.C. Elston(1) (1)Dept. of Epi & Biostat, Case Western Reserve Univ., USA , (2)Dept. of Applied Math and Stat, SUNY at Stony Brook, USA, (3)Lab. of Statistical Genetics, Rockefeller Univ, USA We propose a strategy for maximizing multipoint lod scores in the presence of heterogeneity (HLOD) based on Smith's admixture model. Using the S.A.G.E. program MLOD, the maximization is over the genetic parameter space (allele frequencies, penetrances and admixture proportion) constrained by the prevalence estimated from the corresponding sample. We simulated disease traits under different genetic models on the 30 CEPH families for which complete genomic scan data, 382 markers on 22 autosomes, are available. The trait data are simulated independent of the marker data in order to determine the null empirical distribution of the maximum HLOD scores, and hence cut-offs necessary to control genomewide Type I error.