Code for VNS transcriptomic bioinformatics analysis
Description
Supplementary Data S7 — R scripts for integrative transcriptomic analysis of VNS Content: R scripts and result tables reproducing all transcriptomic analyses of the manuscript; only public RNA-seq data were reanalysed, so no new data or ethics approval were required. Hypothesis: VNS induces a hippocampal-predominant neurodevelopmental transcriptional programme that is brain-region specific (hippocampus yes, cortex no) and partly reproducible across pathological states and in human cells. Data: GSE134309 (rat hippocampus/cortex, implantable VNS 30 min/d x 4 d; hippocampus 3 vs 4, cortex 4 vs 4); GSE246344 (rat hippocampus after MCAO, taVNS; 3 control, 3 stroke, 3 stroke+taVNS); GSE210043 (human PBMC, drug-refractory epilepsy; 4 vs 4). Series matrices from GEOquery v2.68.0, count matrices from GEO Supplementary files; annotation org.Rn.eg.db and org.Hs.eg.db v3.18.0. Scripts: DESeq2 v1.42.0 (mean count <10 removed, negative-binomial dispersion, Wald test) for GSE134309/246344; limma v3.58.0 (linear model, empirical Bayes) for GSE210043; DE thresholds padj<0.1 and |log2FC|>0.5. GSEA: clusterProfiler v4.10.0, genes pre-ranked by log2FC, gseGO on GO BP, p.adjust<0.05 (rat) and <0.25 with top 15 pathways (PBMC). Region specificity: identical DESeq2 pipeline on cortex, comparing the 18 hippocampal developmental genes. Cross-dataset: GSE134309 and GSE246344 GSEA results matched by GO Description, NES quadrants (third concordant, second opposite). Findings: VNS altered [n] hippocampal genes ([n] up, [n] down) versus [n] in cortex; GSEA gave [n] GO BP terms, mainly [terms]; [n] of the 18 genes showed no cortical response; in stroke taVNS changed [n] genes, with [n] concordant and [n] opposite pathways; human PBMC changes were mild ([terms]). Interpretation: log2FC gives direction (positive meaning as defined in each script); NES>0 indicates activation. Since datasets differ in species, tissue, library, depth and normalisation, and batch correction was not possible, compare directions and pathways rather than absolute values; small n (3-4/group) makes GSE210043 exploratory; gene sets are limited to GO BP. Run the scripts in order with R 4.6.0 and the stated package versions after downloading the files listed at the top of each script.