Our investigation of gene–gene/protein–protein interactions highlighted multiple interactions of GWAS-relevant genes with other genome-wide relevant loci or other genes involved in the same molecular pathways. Genetic variations among these interacting partners may have an effect on the predisposition to drug dependence. However, we also found that certain genome-wide relevant genes were not integrated into the network. This suggests that there is missing information in the experimental evidence pertaining to gene–gene/protein–protein interaction in the predisposition to drug dependence – a result that is hardly surprising. Other approaches have been used to integrate drug dependence GWAS data with the human protein interactome [37]. Comparing these approaches, we observed some common features and some differences, which are likely attributable to the fact that the present network analysis is based only on experimental evidence independent from genetic data. Therefore, by analyzing our interaction networks, we have confidently investigated human variation in genes that directly interact with the identified GWAS-relevant genes.
