stLearn: integrating spatial location, tissue morphology and gene expression to find cell types, cell-cell interactions and spatial trajectories within undissociated tissues

Abstract

Spatial Transcriptomics is an emerging technology that adds spatial dimensionality and tissue morphology to the genome-wide transcriptional profile of cells in an undissociated tissue. Integrating these three types of data creates a vast potential for deciphering novel biology of cell types in their native morphological context. Here we developed innovative integrative analysis approaches to utilise all three data types to first find cell types, then reconstruct cell type evolution within a tissue, and search for tissue regions with high cell-to-cell interactions. First, for normalisation of gene expression, we compute a distance measure using morphological similarity and neighbourhood smoothing. The normalised data is then used to find clusters that represent transcriptional profiles of specific cell types and cellular phenotypes. Clusters are further sub-clustered if cells are spatially separated. Analysing anatomical regions in three mouse brain sections and 12 human brain datasets, we found the spatial clustering method more accurate and sensitive than other methods. Second, we introduce a method to calculate transcriptional states by pseudo-space-time (PST) distance. PST distance is a function of physical distance (spatial distance) and gene expression distance (pseudotime distance) to estimate the pairwise similarity between transcriptional profiles among cells within a tissue. We reconstruct spatial transition gradients within and between cell types that are connected locally within a cluster, or globally between clusters, by a directed minimum spanning tree optimisation approach for PST distance. The PST algorithm could model spatial transition from non-invasive to invasive cells within a breast cancer dataset. Third, we utilise spatial information and gene expression profiles to identify locations in the tissue where there is both high ligand-receptor interaction activity and diverse cell type co-localisation. These tissue locations are predicted to be hotspots where cell-cell interactions are more likely to occur. We detected tissue regions and ligand-receptor pairs significantly enriched compared to background distribution across a breast cancer tissue. Together, these three algorithms, implemented in a comprehensive Python software stLearn, allow for the elucidation of biological processes within healthy and diseased tissues. 14/18

Spatially resolved transcriptome profiles of mammalian kidneys illustrate the molecular complexity of functional nephron segments, cell-to-cell interactions and genetic variants

“…From here we focused on the human ST-seq data to investigate cell types, their transcriptional signatures, and spatial locations, using two complementary analytical strategies -Seurat [38] and stLearn [39]. We initially defined the spatial organisation of the human kidney using Seurat and stLearn clustering to identify ST-spots with distinct transcriptome profiles and mapped these cluster identities to the H&E tissue images.…”

“…Additionally, we also had to account for the possibility that each human kidney ST-spot We extended our human kidney ST-seq data analysis to explore cellular communication between glomerular cells using a CCI algorithm [39]. Structurally, a glomerulus is a tuft of capillaries composed of mesangial, endothelial and podocyte cells, surrounded by the Bowman's capsule lined with parietal epithelial cells [3,44].…”

“…Next, we tested >1000 L-R pairs curated in the CellphoneDB database [46,47]. By applying stLearn [39], we mapped L-R gene spatial co-expression within and between ST-spots of glomeruli regions, identifying . CC-BY-NC-ND 4.0 International license perpetuity.…”

“…The generated human and mouse ST-seq data was also analysed using stLearn, a novel Python-based toolkit [39]. stLearn uses the morphological similarity between neighbouring STspots to normalise gene expression and reduce "dropout" noise, an inherent technical limitation of RNA-seq technologies [39,56,57]. With the mouse kidney ST-seq data, we first filtered out genes that were expressed in less than 3 ST-spots.…”

BayesSpace enables the robust characterization of spatial gene expression architecture in tissue sections at increased resolution

“…This approach was later adapted for use with high-throughput spatial transcriptomics data through the selection of spatially differentially expressed genes prior to clustering 12 . Another recently developed spatial clustering algorithm is stLearn, which uses deep learning features extracted from the histopathological images as well as the expression of neighboring spots to spatially smooth the data 13 .…”