Data were screened for differences of each treatment (Ag-MNP, sAg-MNP, and AgNO3) from control using one-way ANOVA with contrasts. The false discovery rate (FDR) threshold was set at 0.2 and the fold change was required to be greater than 1.5, and a P-value <= 0.05. The FDR of 0.2 was selected to balance the protection against false positives while minimizing the rate of false negatives. Using Partek, genes that were significantly different from control in at least one treatment were then analyzed using agglomerative hierarchical cluster analysis (HCA). HCA is a 2-Pass clustering method; the first pass is a K-means clustering and in the second pass the K-means clusters are joined by agglomerative clustering.