-
Notifications
You must be signed in to change notification settings - Fork 159
Description
Description of the bug
We have observed an interesting behaviour when it comes to the bigWig file generation in both the ChIP-seq as well as CUT&RUN workflow, and I wanted to open this up to your collective input.
The ChIP-seq pipeline runs the following code to generate bigWig files:
bedtools genomecov \
-ibam H3K27me3.mLb.clN.sorted.bam \
-bg \
-scale $SCALE_FACTOR \
-pc \
| sort -T '.' -k1,1 -k2,2n > H3K27me3.bedGraph
This default behaviour creates bigWig files that appear to have a relatively high level of background. When using the option --bga
(instead of --bg
), uncovered bins get assigned values of 0 and are not skipped. This appears to result to lower levels of background, which is more pleasing to the eye in IGV:
I have just discussed this with @FloWuenne and @edmundmiller during the weekly helpdesk, and they agreed that this behaviour should be changed (@JoseEspinosa for visibility).
Very detailed additional findings, inclduing IGV screenshots etc, are availalbe on the nf-core Slack: https://nfcore.slack.com/archives/CFP55NZ5G/p1747062371347069
The current ChIP-seq workflow seems to use this local code:
-bg \\ |
But ideally BEDTOOLS_GENOMECOV
should be changed over to using the nf-core module, together with changing the -bg
to --bga
. Users still wishing to specify--bg
could do so via the ext.args
parameter.
Command used and terminal output
Relevant files
No response
System information
No response