Tn5 insertion site adjustment

[z2e2]

Hi ArchR team,

I am trying to use ArchR to process the pbmc_granulocyte_sorted_10k_atac_fragments.tsv.gz file downloaded from 10x Genomics. In your documentation, you mentioned that you do the +4/-5 bp trick on the input data. However, it seems that cell-ranger has adjusted the insertion site according to their doc. There is no strand information (plus-stranded/minus-stranded fragments) in the fragments file from cell-ranger anyway. I am wondering if ArchR only apply the +4/-5 bp adjustment on BAM input and doesn't adjust the insertion sites for fragments from cell-ranger output?

Best,
Joe

[rcorces]

This is stated in the parameter definition for offset adjustment. Shifting only happens for BAM files

ArchR/R/CreateArrow.R

Lines 47 to 52 in 79953a9

	#' @param offsetPlus The numeric offset to apply to a "+" stranded Tn5 insertion to account for the precise Tn5 binding site.
	#' This parameter only applies to bam file input and it is assumed that fragment files have already been offset which is the standard from 10x output.
	#' See Buenrostro et al. Nature Methods 2013.
	#' @param offsetMinus The numeric offset to apply to a "-" stranded Tn5 insertion to account for the precise Tn5 binding site.
	#' This parameter only applies to bam file input and it is assumed that fragment files have already been offset which is the standard from 10x output.
	#' See Buenrostro et al. Nature Methods 2013.