Scale-free coexpression networks were constructed using the weighted gene coexpression network analysis (WGCNA) package in R 22. WGCNA provides a global perspective, emphasizing the interconnectedness between genes to classify different molecular groupings, rather than focusing on individual genes. Alterations in the molecular networks are due to environmental and genetic variation affecting mechanisms of regulation. The molecular groups in WGCNA are defined as modules formed by densely interconnected genes, and alternatively spliced isoforms, which were discerned using a dynamic tree-cutting algorithm based on hierarchical clustering (minimum module size=100, cutting height=0.99, deepSplit=TRUE) 23. Corresponding transcripts were assigned to unique numeric and color identifiers based on the level of read summarization, emphasizing the strongest pairwise relationships in expression across samples. Module preservation, and reproducibility, for alcoholic and matched control subjects was evaluated according to a Z-summary statistic with permutation 24. Assigned modules were functionally annotated against known ontological categories, and additional features such as predicted drug interactions, using standard biological enrichment tools 25, 26. Representative gene ontology plots were visualized using semantic dimension scatter plots to reduce redundancy in terms of the identified
