GH5 thermostability: domain-verified sequence, label, structural and FoldX dataset with analysis code

Curated data and analysis code for a study of thermal adaptation in family-5 glycoside hydrolases (GH5). Includes 757 domain-verified GH5 sequences with optimal-growth-temperature and enzyme temperature-optimum labels, homology-aware clusters, model predictions, 66 charge-adaptation hotspot positions, a salt-bridge and surface-charge survey of 82 GH5 structures, FoldX stability results, and the GH1/GH10/GH11 comparison sets, together with the complete pipeline to reproduce every figure and reported statistic.