Gene expression analysis

Differentially expressed genes were identified using the edgeR (Robinson, et al. 569 2010) package in R. Genes were considered differentially expressed if upregulation or downregulation was ≥ 4-fold relative to the controls with a False Discovery Rate (FDR) ≤ 0.05. Pairwise comparisons were conducted between the Present Day samples and samples subjected to the other treatments.Gene co-expression networks were reconstructed using rlog-transformed expected counts through WGCNA (Langfelder and Horvath 2008) implemented in R. Hub genes (gene significance > |0.2|; module membership > |0.8|) within a module were identified and used in downstream analyses.Pfam and GO enrichment analyses for DEGs and module members were performed using a script (github.com/fle1/canolab_scripts/blob/master/Pfam_enrichment.R) and the topGO package (Alexa and Rahnenführer 2009) implemented in R. Pfam domains and GO terms with p-value < 0.05 were considered significantly enriched.Blastp top hits for each hub gene were used as input in the KEGG pathway enrichment analysis (score > 0.400 and FDR < 0.01) in STRING v.11 (Mering, et al. 614 2003). Relative expression of sponge gene homologs in RCP 8.5 relative to the Present Day control was computed as the average sum of transcripts per million (TPM).