Single-cell RNA sequencing datasets of murine microglia from four sequencing libraries went through initial quality control to exclude low-quality cells by examining the distribution of the number of unique molecular identifiers (UMI) (>1500), the number of genes detected (>1000), the mitochondrial ratio (<0.04), and the overall complexity (log10(genes/UMI) > 0.8). Using Seurat (4.1.1)67,68 and R (4.2.1), datasets were subjected to hashtag antibody oligos (HTO) separation using Seurat’s “HTODemux” with a 0.95 threshold to separate conditions and genotypes and to remove doublets and negative cells. In this step, the second sequencing library (pool 2) was discarded as the expression of the antibody was not enough to separate conditions. A second quality control was performed, and cells with UMI < 4000 and number of expressed genes <2000 were excluded. Genes expressed in less than ten cells were discarded. To remove the batch effect, Harmony69 integration was performed on Seurat’s log-normalized and scaled data while regressing out cell cycle scores (Supplementary Fig. 5a). In this step, a set of genes that lacked biological relevance to our research and displayed uneven expression patterns (Supplementary
