Pan-genome analysis of the glutathione S-transferase gene family reveals the core and dispensable components of GST evolution in Arachis

Introduction Gene-family identification and comparative analyses have largely relied on a single reference genome and therefore cannot capture presence/absence variation among genomes. Methods The glutathione S-transferase (GST) family was analysed computationally across 20 Arachis genomes, 16 allotetraploid and four from the diploid progenitor lineages, using orthologous gene groups (OGGs) as the unit of analysis. Gene structure, phylogeny, duplication, collinearity, Ka/Ks and expression in tissue and stress transcriptomes were compared among OGG categories. Results The 1,436 GSTs carrying both canonical domains were assigned to 57 OGGs, comprising 18 core OGGs present in all 20 genomes and 39 dispensable OGGs (softcore, shell and line-specific). The family was structurally uniform, and its uniformity declined from the core to the shell OGGs. The OGGs fell into the Tau, Phi, Theta and Lambda subfamilies, 53 of the 57 belonging to Tau. WGD/segmental duplicates predominated in core and softcore OGGs, making up 65.2% of core genes, whereas tandem duplicates predominated among shell genes at 42.0%, and WGD/segmental duplicates made up a far larger share of the GSTs in the tetraploids than in the diploids. Core and softcore OGGs occupied syntenic positions shared with the progenitor diploids, about half of the cross-genome syntenic links connecting the A and B subgenomes. Among 23,399 within-OGG gene pairs, mean Ka/Ks was below 1 in 43 of the 46 OGGs for which it could be estimated, consistent with predominantly purifying selection; no pair with Ka/Ks above 1 departed significantly from neutrality, and neither Ka, Ks nor Ka/Ks differed between core and dispensable OGGs. Across 22 tissues of the cultivar NDH108, all 35 core genes were expressed, the nine silent genes belonged to softcore or shell OGGs, and the calyx was the most frequent tissue of peak expression. Under drought, cold or salt, 74 of the 84 NDH108 GSTs responded, at similar rates in core and dispensable OGGs. Discussion All results are computational and were not validated experimentally. The OGG-based analysis resolves variation in GST content and genomic organisation across the Arachis panel that a single reference genome does not capture, and the genes it highlights are candidates for experimental study.