We used principle component analysis (PCA) to analyze gene expression values. The amygdala, PFC and liver samples clustered separately, indicating that tissue region was the largest source of variance in the data (Figure S1A). Hierarchical clustering (HCL) confirmed the PCA results (Figure S1B). Liver samples clustered into two groups: one containing beza and saline samples (not effective for reducing alcohol consumption) and the other containing tesa and feno samples (effective for reducing alcohol consumption) (Figure S1C).
