We correlated transcriptomic data from the central nervous system (CNS) and eyes of male two-toned pygmy squid (Idiosepius pygmaeus) with CO2 treatment levels and OA-affected behaviours in the same individuals to identify genes potentially contributing to OA-induced behavioural changes. We used Weighted Gene Co-expression Network Analysis (WGCNA) to cluster transcriptome-wide gene expression into modules. The gene expression profile of each module was then correlated with CO2 treatment levels (current-day: ~450 µatm, elevated: ~1,000 µatm) and visually-mediated behavioural responses previously shown to be altered at elevated CO2, using Canonical Correlation Analysis.
All statistical analyses were carried out in R (v4.0.4), primarily using RStudio (v 1.4.1106).
DESeq2 (v1.30.1) was used to normalise gene count data, remove low read counts and for variance stabilisation.
WGCNA (v1.70-3) was used for network construction and module detection.
CCA (v1.2.1) was used to explore the correlations between the two sets of variables from the same individual squid: set 1 = module eigengenes from each module representing the gene expression profile of the module, set 2 = CO2 level (current-day or elevated) and behavioural traits.
WGCNA (v1.70-3) was used to calculate module membership and gene significance to identify hub genes. clusterProfiler (v3.18.1) was used to run over-representation analysis using the hypergeometric test.
This data record contains:
1) All R code used for the statistical analyses (.html file)
2) Data files to accompany the statistical analyses = gene count data for the CNS and eyes of each individual squid (corset-counts.txt), CO2 treatment level and behavioural measurements for each individual squid (metadata.csv), explanation of each variable in metadata.csv (README.txt), annotated transcriptome assembly of Idiosepius pygmaeus CNS and eye tissues (annotated_transcriptome.csv), data linking each transcript from the transcriptome assembly to each gene in the gene count data (corset-clusters.txt)
Software/equipment used to manipulate/analyse the data: DESeq2 (v1.30.1)
WGCNA (v1.70-3)
CCA (v1.2.1)
clusterProfiler (v3.18.1)
Majority of analyses: Platform: x86_64-w64-mingw32/x64 (64-bit) Running under: Windows 10 x64 (build 18363)
WGCNA network construction and module detection: Platform: x86_64-pc-linux-gnu (64-bit) Running under: CentOS Linux 8 (Core)
The HTML file contains all R code used for the analyses, as well as the Session Information.