Source data for "Accurate trajectory inference in time-series spatial transcriptomics with structurally-constrained optimal transport"

These files are used to generate several files in our publication, "Accurate trajectory inference in time-series spatial transcriptomics with structurally-constrained optimal transport."The file "synthetic_data" contains several synthetic datasets we generated to evaluate our method, Spatiotemporal Optimal transport with Coherent Structures (SOCS)--one t1 dataset and several t2 datasets, representing ST data with a single structure translated by different amounts.Each folder contains the data for one figure panel, specified by the folder name.Within each folder is contained .csv files giving the spatial coordinates and information about structure assignment or cell type assignment for the cells illustrated in the specified panel.Additionally, each folder contains one or more .npy (numpy) file, representing a learned transport map, an array in which element (i,j) gives the mass of cell i at time point 1 mapping to cell j at time point 2.Row i in the transport map file corresponds to row i in the t1 metadata file, and column j in the transport map corresponds to row j in the t2 metadata file.