Using genetic information in risk prediction for alcohol dependence

摘要: USING GENETIC INFORMATION IN RISK PREDICTION FOR ALCOHOL DEPENDENCE By Jia Yan, B.A. A dissertation submitted in partial fulfillment of the requirements for degree Doctor Philosophy Human and Molecular Genetics Master Science Genetic Counseling at Virginia Commonwealth University. University, 2012. Major Director: Danielle M. Dick, PhD Associate Professor Psychiatry, Psychology, Family-based genome-wide association studies (GWAS) alcohol dependence (AD) have reported numerous associated variants. The clinical validity these variants predicting AD compared to family history has not yet been reported. These aim explore aggregate impact multiple genetic with small effect sizes on risk prediction order provide a interpretation contributions AD. Data simulations showed that given AD’s prevalence heritability, model incorporating all would an area under receiver operating characteristic curve (AUC) approaching 0.80, which is often target AUC screening. Adding additional environmental factors could increase 0.95. Using Collaborative Study Alcoholism (COGA) Addiction: Genes Environment (SAGE) GWAS samples, we used several different sources capture information discovery then tested sum scores created based this predictive accuracy validation ii samples. Scores were assessed separately single nucleotide polymorphisms (SNPs) candidate gene analyses. Candidate did exhibit significant accuracy, but SNPs meeting less stringent p-value thresholds analyses did, ranging from mean estimates 0.549 p<0.01 0.565 p<0.50. Variants subtypes there similarly modest ability externalizing subtype. individual SNP effects across entire genome accounted 0.46%-0.57% variance symptom count, AUCs 0.527 0.559. Additional covariates are correlated increased 0.865. Family was better classifier case-control status than scores, 0.686 COGA 0.614 SAGE. This project suggests currently limited validity, potential enhanced detection contributing