HistoGWAS-dataset

This dataset contains histology image embeddings derived from whole-slide images across 11 human tissues analyzed in the HistoGWAS study. Whole-slide images were tiled into patches at 192 µm spatial resolution, and patch-level embeddings were extracted using a pretrained RetCCL encoder. The embeddings were used for downstream genetic association analyses and visualization of genotype-associated phenotypic variation. The dataset also includes corresponding highly variable gene expression matrices for the same tissues to enable integrated multimodal analyses of tissue morphology and gene expression.