Picard CollectHSMetrics: how do mean_coverage; max_coverage and min_coverage get calculated when using --PER_TARGET_COVERAGE option?
Hi there!
I am looking into Picard CollectHSMetrics for some hybridization-based protocol, but having some difficulties interpreting the metrics computed by Picard. when running the --PER_TARGET_COVERAGE option. Could someone advice on how to interpret the output of this (specifically how mean_coverage, max_coverage and min_coverage are computed), please?
Thank you so much for your help!
Monica
Some details below:
Picard version used: 2.27.5. Run the following command
picard CollectHsMetrics -I {input.bam} -O {sample}.collectshsmetrics_raw.txt -R {REF} -PER_TARGET_COVERAGE {sample}.PerTargetCov.txt -BAIT_INTERVALS {config[INTERVALS]} -TARGET_INTERVALS {config[INTERVALS]} -COVERAGE_CAP 1000 -NEAR_DISTANCE 500
-
--PER_TARGET_COVERAGE option generates an addtional file to display minimum, mean and median per nucleotide coverages per capture target provided to the tool. Tool selects and filters reads based on Mapping Quality, Base Quality and overlapping reads are counted out for the coverage calculations.
I hope this helps.
-
Hi Gökalp Çelik,
That's very helpful - thank you!
Would you then say that by doing sum(read count)/total reads, I could have an estimation of the % of reads that overlap the targeted regions supplied (i.e. % of on-target reads)?
Thanks so much again!
-
Yes. For more detailed descriptions you may check the table in the page below.
https://broadinstitute.github.io/picard/picard-metric-definitions.html#HsMetrics
-
Thank you for your help!
Please sign in to leave a comment.
4 comments