This is a test version of Biostars. For the public version, visit https://www.biostars.org.
Difference between samtools unmapped reads and picard metrics file unmapped reads

I have noticed that the count of unmapped reads from:

samtools view -f 4 $sample.sam > $sample.unmapped.sam

is LESS than (about ~90% of) what I got from the "UNMAPPED_READS" column in the metrics file generated by MarkDuplicates via picard:

java -Xmx45g -jar /dir/picard.jar MarkDuplicates INPUT=$sample.sam OUTPUT=$sample.sort.mdup.bam METRICS_FILE=$sample.metrics.mdup.txt

I checked picard definition of UNMAPPED_READS, and it is: "The total number of unmapped reads examined. (Primary, non-supplemental)". I don't understand why the two numbers do not match. Any suggestion please? Thank you.

samtools picard
• 148 views • link
updated by Hussain Ather Hussain Ather • written by DVA DVA

1 answer

Do all of the reads have a MAPQ of 0 for both the Picard and samtools output?

• 0 views • link
written by Hussain Ather Hussain Ather

I do not have an unmapped reads output from Picard - I do not know how to do it in Picard, I was just looking at the metrics file. That said, are you implying that in Picard, MAPQ = 0 is counted as unmapped? I think I can write a script to prove that. Thanks a lot for the reply.

• 0 views • link
written by DVA DVA

Log in to answer this question.

More posts like this