Question

Question on removing duplicates with ATAC-Seq and ChIP-Seq

1

Entering edit mode

12 months ago

tamu.anand ▴ 30

Hi

I am new to ATAC-Seq and ChIP-Seq and I have a question on removing duplicates when it comes to paired-end ATAC-Seq and paired-end ChIP-Seq pipelines.

I have seen some papers where clumpify is used in the first step to remove duplicates followed by bwa-mem and passing the bam files (after shifting for ATAC-Seq) to MACS2 with no usage of Picard MarkDuplicates.

https://www.cell.com/cell-genomics/pdf/S2666-979X(23)00019-8.pdf
Single nucleus multiomics identifies ZEB1 and MAFB as candidate regulators of Alzheimer’s disease-specific cis-regulatory elements
ChIP-seq analysis
Prior to analysis, reads were processed to remove optical duplicates with clumpify (BBMap v38.20; https://sourceforge.net/projects/
bbmap/) [dedupe = t optical = t dupedist = 2500]

I have seen this post - Did you remove ChIP-seq duplicates - where Picard MarkDuplicates is used

Hence just curious - is there benefits with one approach or the other (Approach 1 - using clumpify as step 1 and no MarkDuplicates; Approach 2 - use aligned bam files and then do MarkDuplicates)

If using clumpify, is the above correct - dedupe=t optical=t dupedist=2500

Removing fastq duplicates GenoMax suggests hdist=0 for making it strict
Yes .. BBMap can do that! - Part III clumpify (mark (and dedupe) duplicates without alignment), mutate (create mutant genomes) and other miscellaneous tools should I use dedupe=t optical=f or dedupe=t optical=t

I would like to seek your advice and guidance on the above for both ATAC-Seq and ChiP-Seq.

Would anything change if it was single-end instead of paired-end as alluded to here: Did you remove ChIP-seq duplicates

Thanks in advance.

clumpify chip-seq atac-seq bbmap • 823 views

ADD COMMENT • link 11 months ago by tamu.anand ▴ 30

1

Entering edit mode

Using clumpify.sh will allow you to do duplicate removal in an alignment independent manner. It will also allow you to remove just optical (really clustering) duplicates, which is what the paper you linked above seems to be doing. It will also make the size of the input files smaller thus reducing time required for alignment.

Keep this statement from @Ian sudbery's answer linked above in mind

If you are using MACS for your peak-calling, you'll want to mark duplicates rather than remove them.