Extended Data Fig. 1: ATAC-seq quality control (QC) and OCR peak calling.
From: PICALM Alzheimer’s risk allele causes aberrant lipid droplets in microglia

(a) Representative images of IF staining (left) of iMG (TREM2+CD45+PU1+) used for ATAC-seq and iMG purity quantification (right). Each datapoint represents one single-well measurement from one experiment of one donor line (n = 5). Scale bar, 50 μm. (b) Representative images of IF staining (left) of iAst (S100B+, VIM+, or GFAP+) used for ATAC-seq and iAst purity quantification (right). Note that GFAP staining and imaging was performed independently from S100B/VIM staining and imaging. Each datapoint represents one single-well measurement from one experiment of one donor line (n = 5 lines). Scale bar: 50 μm. (c) ATAC-seq fragment size distribution plots show periodical nucleosome-free region patterns that are similar between samples and different iPSC-derived cell types. (d) and (e) show comparable and acceptable according to the ENCODE ATAC-seq guideline version 4: FRiP (fragment reads in peaks) score > 0.3, TSS (transcription start site) annotation > 5 across cell types. (f) iMG-specific OCR peaks for a known MG-specific gene SPI1 (also known as PU.1) and much stronger peaks for VIM in iAs than in other cell types. (g) PCA clustering of human iPSC-derived MG, Ast, NGN2-Glut, GABAergic, and dopaminergic neurons. Normalized ATAC-seq reads derived from a 501-bp union peak set from all five cell types (622,987 peaks in total) were used for PCA. (h) UpSet plot showing the cell type-specific ATAC-seq peaks and shared peaks across different cell types. (i) PCA of ATAC-seq peak accessibility of iMG and iAst samples from the current study in comparison to previously reported datasets of human brain MG and Ast (circled areas) or iAst as well as other brain cell types6,22. 210,833 peaks from the previous studies6,22 were used for PCA. GABA, GABAergic neurons; hOlig, human oligodendrocytes; MGAS, the mixture of microglia and astrocytes.