Analysis of chromatin datasets can be highly dependent on accurate peak calling and this challenge is compounded in single-cell assays where peaks specific to rare populations are sometimes missed when calling peaks on the whole cell population. To address this problem, we identified peaks using MACS2 (ref. 46) for each annotated cell type separately and combined the individual peak calls into a unified peak set using Signac. Indeed, peaks specific to rare cell populations were often missed when calling peaks on the whole dataset using MACS2 (Supplementary Fig. 1a,b). We further compared the MACS2 bulk-cell peak calls with the peak calls produced by 10x Cellranger ATAC v.1, commonly used for the analysis of scATAC-seq data and found 13,751 cases where a Cellranger peak merged distinct MACS2 peaks into a single region, whereas there were only two cases where a MACS2 peak overlapped multiple Cellranger peaks (Supplementary Fig. 1c). This revealed a bias in Cellranger for aberrant merging of multiple distinct peaks into a single region and highlights the importance of accurate cell-type-specific peak calling methods in the analysis of single-cell
