CellPie: a scalable spatial transcriptomics factor discovery method via joint non-negative matrix factorization

Spatially resolved transcriptomics has enabled the study of expression of genes within tissues while retaining their spatial identity. Most spatial transcriptomics technologies generate a matched histopathological image as part of the standard pipeline, providing morphological information that can complement the transcriptomics data. Here we present CellPie, a fast, unsupervised factor discovery method, based on joint non-negative matrix factorisation of spatial RNA transcripts and histological image features.CellPie employs the accelerated hierarchical least squares method to significantly reduce the computational time, enabling efficient application to high-dimensional spatial transcriptomics datasets. We assessed CellPie on two different human cancer types and spatial resolutions, showing an improved performance against published factorisation methods. Additionally, we applied CellPie to a highly resolved Visium HD dataset, demonstrating its high computational efficiency compared to standard non-negative matrix factorisation and other existing methods. Availabilityhttps://github.com/ManchesterBioinference/CellPie