To evaluate the impact of Cd83 knockout on cellular composition during EAE, we employed MELD (version 1.0.0, implemented in Python 3.8.8)38. PHATE dimension reduction was performed, followed by PCA. By running “meld” with default settings, we computed densities associated with replicates. The resulting densities were normalized (using L1 normalization to ensure that the sum of each cell’s values equals 1) and averaged across all replicates for each condition per cell. Condition-associated relative likelihood in each graph-based cluster was visualized. Next, vertex frequency clustering (VFC) was performed on merged MHCII+, proliferating and Cxcl10+ clusters, and each major cluster was divided into subclusters based on the visual distribution of the EAE-wildtype associated likelihood.
