Package: spicyR 1.99.11

spicyR: Test Changes in the Co-Localisation of Cell Types Between Groups of Patients

spicyR tests whether the spatial co-localisation of pairs of cell types differs between groups of patients, or is associated with patient outcome, in cell-resolution spatial omics data (imaging mass cytometry, CODEX, MIBI, Xenium, CosMx and others). Its default method, spicyR Cell, counts the cells of one type within a radius of each cell of another and compares the count with its exact expectation under random labelling of the observed cells. Images are combined within patients by a frailty generalised estimating equation, and differences are tested with a small-sample cluster-robust (CR2) variance with patients as the units. Covariates, several conditions, several radii and survival outcomes are supported, and the original image-level test is kept as an option. The computations run in a C++ core shared with the Python package spicyr.

Authors:Nicolas Canete [aut], Ellis Patrick [aut, cre], Sadiq Dohadwalla [aut], Elijah Willie [aut], Nicholas Robertson [ctb], Alex Qin [ctb], Farhan Ameen [ctb], Shreya Rao [ctb]