SpaCeNet: Spatial Cellular Networks from Omics Data
摘要
We propose SpaCeNet as a method for analyzing patterns of correlation in spatial transcriptomics data. SpaCeNet extends the concept of conditional independence to spatially distributed information, and disentangles conditional independence relations between inter- and intra-cellular variables. Moreover, SpaCeNet addresses the various length scales over which intercellular communication occurs via flexible interaction potentials in combination with appropriate regularization strategies.