The question about peak calling

[superSXB]

During the analysis of Signac, can peaks be re-called after importing cell types? If some of my peaks are contributed by specific cell subtypes, their signals may be weaker than the background in the overall peak set, and this information may be lost in the first step of peak calling by Cell Ranger. Is it possible to obtain cell types and then call peaks for each cell type and finally merge them to form the overall peak set?