Giters

jianhong / ATACseqQC

ATAC-seq Quality Control

Home Page:https://jianhong.github.io/ATACseqQC/articles/ATACseqQC.html

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

adjusting readBamFile output for plotCorrelation

tinavisnovska opened this issue · comments

commented

Hi,
(in ATACseqQC 1.12.0)
the example of using plotCorrelation that is written in the ATACseqQC guide works out of the box only when asMates and bigFile parameters are FALSE in readBamFile. Otherwise output format of readBamFile does not cooperate with plotCorrelation, is there a way around to modify output of readBamFile with asMates/bigFile being TRUE so that plotCorrelation can be used? And more importantly, does it make sense to use plotCorrelation in such cases?
Thanks for any help!

commented

Thank you for reporting that. I will fix it as soon as possible.

commented

Hello, JianHong!
Thanks for the excellent tool!
 
However, I met some issues while running the ATACseqQC. One of them is to perform the plotCorrelation analysis.

  1. Could you please tell me which bam file I should use for this analysis. The raw one, the clean one after bamQC step, or the one after shiftGAlignmentsList step? Besides, since ATACseqQC requires huge resources and lots of time for big BAM, I perform the analysis function by function (not run all of them at one time). I am confused by which bam I should use in each step as well. I would appreciate if you can explain it for me.
  2. I ran plotCorrelation analysis with three big bams which contain PE alignment, but it fails. I have no idea if the problem above has been fixed or remained what it was as two years ago?
  3. The main code and error information was listed as follows.
gals <- lapply(bamfiles, function(bamfile){readBamFile(bamFile=bamfile, tag=tags,which=which, bigFile=TRUE,asMates=TRUE)})

plotCorrelation(GAlignmentsList(gals), txs, seqlev=allseq)
Error in GAlignmentsList(gals) :
  all elements in '...' must be GAlignments objects

Could you please have a look at it? Many thanks in advance!

commented

Please refer the documentation of plotCorrelation. I add a sentence to address this issue:

The correlation will be calculated by the correlation of insertion sites within promoter regions. Even
the sequencing is paired-end, please treat it as single ends.
commented

Thank you so much. It works!