Biostar Beta. Not for public use.
Question: shall we consider -- keep-dup all when we removed duplication?
0
Entering edit mode

I have done alignment and then remove duplication with both samtools flag and picard, I like to know if I want to do peak calling using MACS2, should I consider -- keep-dup all or let MACS2 remove duplicate tags one more time?

ADD COMMENTlink 12 months ago star • 150 • updated 12 months ago benformatics • 870
2
Entering edit mode

Unless you merged different libraries in a downstream step (post-duplicate removal) or alternative deduplicated using UMIs (or something similar) then it shouldn't matter because you should no longer have duplicate reads in your samples.

Feel free to try both with and without; the results should be identical.

ADD COMMENTlink 12 months ago benformatics • 870
Entering edit mode
1

It depends on if he really removed duplicated reads, or just marked them.

star , could you show the commands you used?

ADD REPLYlink 12 months ago
h.mon
25k
Entering edit mode
1

Yes that would be true but hopefully when he/says that says they remove them and mentions samtools flags AND picard. My sincerest hope is that they flagged the duplicates with samtools and removed with MarkDuplicates (parameter: REMOVE_DUPLICATES=true)

ADD REPLYlink 12 months ago
benformatics
• 870
Entering edit mode
1

Thank for reply. Yes I did it with bouth and when I have checked with 'samtools flagstat' there was no duplicate. Also I checked with both --keep-DUP all and default command , and there was no difference between number of reads/ peaks.

ADD REPLYlink 12 months ago
star
• 150

Login before adding your answer.

Similar Posts
Loading Similar Posts
Powered by the version 2.0