Picard Markduplicates Problem

Question

[xd_d]

Hello,

here is my problem ! I run picard MarkDuplicates with REMOVE_DUPLICATES=true. And I thought that remove pcr and optical duplicates. But it doesn't work. I dont know what the problem is. But this are duplicates or ? samtools rmdup remove these duplicates.. It seems that picard cluster the duplicates but don't remove these duplicates..

Is this a bug or what is wrong ?

FCC0WRYACXX:1:1212:8677:81169   99  1   14405   1   100M    =   14489   183 TTCTGCTCAGTTCTTTATTGATTGGTGTGCCGTTTTCTCTGGAAGCCTCTTAAGAACACAGTGGCGCAGGCTGGGTGGAGCCGTCCCCCCATGGAGCACA    @CCFFFFFHHHHDGIIJJJGCDJGDEEHIDGHCAGGIIEFHHCFHHCFHIGHIIJGGEHEHGHJIIGEAB6BCCDCB;<BCBB?BDBDDDDDCC>ACBBD    PG:Z:MarkDuplicates NH:i:4  HI:i:1  nM:i:0  AS:i:197

FCC0WRYACXX:1:1213:5301:23125   419 1   14405   1   100M    =   14492   187 TTCTGCTCAGTTCTTTATTGATTGGTGTGCCGTTTTCTCTGGAAGCCTCTTAAGAACACAGTGGCGCAGGCTGGGTGGAGCCGTCCCCCCATGGAGCACA    @BCFDEFFHHFHFIIJIIJJIIIGGFHHHGIEFHGJJIJGGGGIGHHGJJEFFHEGIGHHIFCFGGIHHFFFEDC5;?BDD@DBDDDDDBB?9>CB?BBC    PG:Z:MarkDuplicates NH:i:4  HI:i:3  nM:i:0  AS:i:198

FCC0WRYACXX:1:1308:4376:65126   99  1   14405   1   100M    =   14491   186 TTCTGCTCAGTTCTTTATTGATTGGTGTGCCGTTTTCTCTGGAAGCCTCTTAAGAACACAGTGGCGCAGGCTGGGTGGAGCCGTCCCCCCATGGAGCCCA    @@@FFFFFHHHHDIHIEFGFHIIIIAHFDHHHHIIJGIIIJIGIGIJJJJJJGIJJJIIIJ@CHGGAEAB<>ABD??B5?CDBB5<>ABD9CCD:<C(<A    PG:Z:MarkDuplicates NH:i:4  HI:i:1  nM:i:1  AS:i:196

[xd_d]

maybe picard flag unmapped duplicates but dont remove them ?

[Brian Bushnell]

If the mate is mapped at a different location, a pair is not a duplicate...

[xd_d]

I had reading that a mate paire is a duplicate if they have the same start position and they have all the same start position

[Brian Bushnell]

The mate position is different, though. So they are not duplicates and you would not want to remove them.

Answer 1

[BioinfGuru]

First: It is better to read the documentation before posting a question

Second: Stick to picard markduplicates for this, not sam remdup

Third: This is what you are looking for...

https://broadinstitute.github.io/picard/command-line-overview.html

"The program can take either coordinate-sorted or query-sorted inputs, however the behavior is slightly different. When the input is coordinate-sorted, unmapped mates of mapped records and supplementary/secondary alignments are not marked as duplicates. However, when the input is query-sorted (actually query-grouped), then unmapped mates and secondary/supplementary reads are not excluded from the duplication test and can be marked as duplicate reads."

Your result is dependent on: 1) how you sort 2) whether you want to keep or remove unmapped mates

I would suggest sorting by coordinate and removing unmapped mates before running picard mark duplicates.

How to do this?..... read about SAM flags and use the command line.