Fig. 2

Spatial tendency analysis finds genes associated with spatial structures. a HE image of the human dorsolateral prefrontal cortex (DLPFC) sample. Regions of white matter (WM) and six cortical layers (L6 to L1) are labeled on the image. b Regression curves of gene clusters between gene expression and the distance to WM. Polynomial regression models were fitted to identify genes whose expression varied along with the distance to WM boundary. These genes were grouped into 10 clusters by K-means clustering algorithm. Each curve presents a cluster of genes with similar spatial expression tendency. Zero at the horizontal axis indicates the outer boundary of WM. c Association between gene clusters and previously reported layer specific genes. Each row corresponds to a prior gene-list that specifically expresses in one neuronal layer [30]. Each red unit indicates the cluster of genes (column) is enriched in the prior gene-list (row). d Spatial distributions (top) and fitted curves (bottom) of the representative genes