Aim and Objective: The disturbance of consciousness is one of the most common
symptoms of those have alcoholism and may cause disability and mortality. Previous studies
indicated that several single nucleotide polymorphisms (SNP) increase the susceptibility of
alcoholism. In this study, we utilized the Ensemble Bayesian Network (EBN) method to identify
causal SNPs of alcoholism based on the verified GAW14 data.
Materials and Methods: We built a Bayesian network combining random process and greedy search
by using Genetic Analysis Workshop 14 (GAW14) dataset to establish EBN of SNPs. Then we
predicted the association between SNPs and alcoholism by determining Bayes’ prior probability.
Results and Conclusion: Thirteen out of eighteen SNPs directly connected with alcoholism were
found concordance with potential risk regions of alcoholism in OMIM database. As many SNPs
were found contributing to alteration on gene expression, known as expression quantitative trait loci
(eQTLs), we further sought to identify chemical compounds acting as regulators of alcoholism genes
captured by causal SNPs. Chloroprene and valproic acid were identified as the expression regulators
for genes C11orf66 and SALL3 which were captured by alcoholism SNPs, respectively.