Question: Conflict in cnvr and cnvs result from cn.mops
0
gravatar for thestaroceanster
2.6 years ago by
thestaroceanster0 wrote:

So currently I have been using cn.mops for CNV detection. The result for one sample is confusing since there is overlap between different CNV regions in cnvr output.

It reads as follows:

  seqnames start end width strand X1315_1.final.bam
1 1 144813741 249213345 104399605 * CN3
2 2 13140524 13150289 9766 * CN2
3 3 44143016 44170147 27132 * CN2
4 4 19526785 19545625 18841 * CN3
5 7 69864128 69924740 60613 * CN2
6 7 73050706 124472685 51421980 * CN4
7 7 86116372 115099423 28983052 * CN2
8 8 19904258 59107586 39203329 * CN2

 

It appeared to be an overlapping in chr7 with [86116372,115099423] assigned with CN2 while in the previous line it was given CN4. It's confusing, so I re-check the cnvs output.

 

It reads:

 

21 7 38217807 38937729 719923 * 1315_1.final.bam -1.000003911 -2.0535 CN1
22 7 56496077 56949839 453763 * 1315_1.final.bam -0.472927838 -1.7989 CN1
23 7 121946569 121950131 3563 * 1315_1.final.bam 5.887672454 4.6284 CN4

 

It's not overlapping however it doesn't exactly consistent with cnvr output.

 

SO the question is: which output should I use? Is there an explanation for the difference between cnvr and cnvs? what may be the cause of the cnvr results overlapping?

Here is the command I use:

library(cn.mops)

BAMFiles <- c('/home/export/data/P101SC17020317-01juyongzhi/Results/Bam/1315_1.final.bam','/home/export/data/P101SC17020317-01juyongzhi/Results/Bam/1315_2.final.bam','/home/export/data/P101SC17020317-01juyongzhi/Results/Bam/sq1315_pbmc.final.bam')

 

segments <- read.table('/home/export/data/1194/bam/result_bed_no_chr.txt',sep="\t",as.is=TRUE)

segments <- unique(segments)

gr <- GRanges(segments[,1],IRanges(segments[,2],segments[,3]))

X <- getSegmentReadCountsFromBAM(BAMFiles,GR=gr)    

resRef <- referencecn.mops(cases=X[,1],controls=X[,3],classes=c("CN0", "CN1", "CN2", "CN3", "CN4", "CN5", "CN6","CN7","CN8","CN16","CN32","CN64","CN128"),I = c(0.025, 0.5, 1, 1.5, 2, 2.5, 3, 3.5, 4, 8, 16, 32, 64),segAlgorithm="DNAcopy")

resRef <- calcIntegerCopyNumbers(resRef)

segm <- as.data.frame(segmentation(resRef))
CNVs <- as.data.frame(cnvs(resRef))
CNVRegions <- as.data.frame(cnvr(resRef))
    write.csv(segm,file="/home/export/CG/copynumber_clonal_evolution/result/1315/1315_1_segmentation.csv") 
    write.csv(CNVs,file="/home/export/CG/copynumber_clonal_evolution/result/1315/1315_1_cnvs.csv")
    write.csv(CNVRegions,file="/home/export/CG/copynumber_clonal_evolution/result/1315/1315_1_cnvr.csv")

cn.mops • 607 views
ADD COMMENTlink modified 2.5 years ago by Günter Klambauer540 • written 2.6 years ago by thestaroceanster0
Answer: Conflict in cnvr and cnvs result from cn.mops
0
gravatar for Günter Klambauer
2.5 years ago by
Austria
Günter Klambauer540 wrote:

Sorry for the late response - I thought we clarified this via email.

The segments in the CNVR slot are a union of the individual CNVs. For example, if individual A has a CNV in segments 3 and 4, and individual B has a CNV in segment 4 and 5, the CNVR-slot will contain a CNV-region spanning the three segments [3,4,5].

So, if you are looking for CNVs in individuals with accurate breakpoints, you should use the "CNVS"-slot. If you are looking for regions with high genetic variance (e.g. for GWAS-studies to decrease FDR and increase discovery power), you can use the CNV regions (CNVR).

Regards,

Günter

ADD COMMENTlink written 2.5 years ago by Günter Klambauer540

Thanks for the answer. However, I only use cn.mops for only one sample (1315_1). So there should not be a second individual.

ADD REPLYlink written 2.5 years ago by thestaroceanster0
Please log in to add an answer.

Help
Access

Use of this site constitutes acceptance of our User Agreement and Privacy Policy.
Powered by Biostar version 16.09
Traffic: 336 users visited in the last hour