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.

Most polygenic risk score (PRS)models have been based on data from populations of European origins (accounting for the majority of the large genomics datasets, e.g. >78% in the UK Biobank and >85% in the GTEx project). Although several large-scale Asian biobanks were initiated (e.g. Japanese, Korean, Han Chinese biobanks), most other Asian countries have little or near-zero genomics data. To implement PRS models for under-represented populations, we explored transfer learning approaches, assuming that information from existing large datasets can compensate for the small sample size that can be feasibly obtained in developing countries, like Vietnam. Here, we benchmark 13 common PRS methods in meta-population strategy (combining individual genotype data from multiple populations) and multi-population strategy (combining summary statistics from multiple populations). Our results highlight the complementarity of different populations and the choice of methods should depend on the target population. Based on these results, we discussed a set of guidelines to help users select the best method for their datasets. We developed a robust and comprehensive software to allow for benchmarking comparisons between methods and proposed a computational framework for improving PRS performance in a dataset with a small sample size. This work is expected to inform the development of genomics applications in under-represented populations. PRSUP framework is available at: https://github.com/BiomedicalMachineLearning/VGP