Question: WES samples failing ABSOLUTE
gravatar for mkaushal
3.2 years ago by
United States
mkaushal110 wrote:

I am trying to determine the purity and ploidy of my WES samples using ABSOLUTE. I have normal, primary and metastasized samples. Some of the samples failed ABSOLUTE. Here is an example from the error file:

[1] "Capping 13 segs at tCR = 5.0"

[1] "Expected copy-ratio = 0.63663"

[1] "Mode flag is NA, not generating plots. Sample has failed ABSOLUTE"

[1] TRUE

When I checked the value of seg.dat[["mode.res"]][["mode.flag"]] in the Rdata file for the failed samples, the reason for failure is stated as "E_CR_SCALE" (the expected copy number was too far away from 1, either <0.75 or >1.25. Check that your seg file is not empty). The segmentation files I am using are generated from Varscan copy number -> DNA copy segmentation programs and they are not empty. These are the raw segmentation results from DNA copy and are not merged or filtered. I am also using maf files with somatic snps called by Mutect.

Can anyone help me understand the reason behind my failed runs. What does capping segs at tCR = 5.0 mean? Is there any parameter that can be adjusted to deal with these samples? Any insight would be great.

purity ploidy wes absolute • 2.1k views
ADD COMMENTlink modified 3.2 years ago by markus.riester500 • written 3.2 years ago by mkaushal110

Hi mkaushal,

Did you solve this problem? I'm facing the same problem as you while using varscan + DNcopy to generate the input file for ABSOLUTE. I also got the "Mode flag is NA" error. I found that you posted many questions about the usage of ABSOLUTE. Hope you have some suggestions!

Thank you! gmy

ADD REPLYlink written 2.5 years ago by tiejiagmy0
gravatar for markus.riester
3.2 years ago by
markus.riester500 wrote:

Looks like the copy number ratios are not in the expected format and range. Check that ABSOLUTE knows that you are providing total copy number ratios and not allele-specific (HAPSEG) ones, i.e. set copy_num_type argument to "total". Make sure that they are correctly log-transformed.

As a general recommendation, ABSOLUTE doesn't work well with DNAcopy-like input. Without germline SNPs, you cannot identify the correct solution among the multiple ones compatible with the coverage ratio data. The provided MAF might help in forcing the correct purity, but that doesn't help with ploidy ambiguity. So I would recommend trying to get HAPSEG to work with your data or try some of the recent tools designed for WES data. Most of them work from BAM files.

ADD COMMENTlink written 3.2 years ago by markus.riester500
Please log in to add an answer.


Use of this site constitutes acceptance of our User Agreement and Privacy Policy.
Powered by Biostar version 2.3.0
Traffic: 895 users visited in the last hour