When we compared the M-cell-positive subset of inflamed samples to healthy samples, though, M cells were indeed credibly increased. In the lamina propria, B-cell subpopulations showed several changes, e.g., a decrease of plasma B cells with disease (validated with stainings in Smillie et al.1), and an increase of follicular B cells. Moreover, consistent with our simulation studies demonstrating scCODA’s higher sensitivity for lowly abundant cell types, scCODA uniquely detected statistically credible changes in several low-abundant immune cell populations. For instance, scCODA identified regulatory T cells (Treg) to be more abundant in UC patients which is consistent with other studies21. Smillie et al. combined the results of their Dirichlet regression with two non-compositional tests, Fisher exact test and Wilcoxon rank-sum test, to identify absolute changes in each population independently. Using such a two-stage procedure, Smillie et al. also reported changes in the low-abundant cell types such as Treg cells. For comparison, ANCOM only identified significant changes in inflammatory fibroblasts (healthy vs inflamed), epithelial cells and pericytes (healthy vs non-inflamed), while all cell types in non-inflamed vs inflamed were reported as significantly changing. In contrast to Dirichlet Regression, scCODA reported credible changes in inflammatory fibroblasts (IAFs) for the healthy vs. inflamed case.
