Re: CODEX2 weird behavior

[Jiang, Yuchao]

There’s no bug. The exonic targets you have are overlapping. They should not.

Yuchao

ref=ranges(ref_qc[seqnames(ref_qc)=='chr14'])

countOverlaps(ref,ref)

On Jan 17, 2019, at 9:27 AM, Janka Puterová <iput...@fit.vutbr.cz> wrote:

Yes, the data comes from the same sample: MM-ARK-07-418-aPC. I am sending you a link to all output data I saved during CODEX2 analysis as well as RData workspaces.

https://www.dropbox.com/s/nhyewjnbiytxo7w/codex_results.tar.gz?dl=0

Please, see below the description of files:

• MM_copynumbers.csv - resulting copynumbers
• MM_coverage.csv - raw read depth (coverageObj)
• MM_qcmat.csv - quality control (qcmat)
• MM_copynumbers.Rdata - contains everything from preprocessing to running codex2 from the demo until codeline N <- apply(apply(Y.nonzero, 2, function(x){x/pseudo.sample}), 2, median) 
• MM_workspace.RData - contain everything from Running CODEX2 with negative control samples to end of analysis

Janka

On 16.01.2019 17:07, Jiang, Yuchao wrote:

Hi Janka,

 

Are these from the same sample? Can you send me the normalized data and raw read count data, as well as the ref_qc object and I will take a look?

 

Yuchao

 

 

From: Janka Puterová <iput...@fit.vutbr.cz> 
Sent: Wednesday, January 16, 2019 9:58 AM
To: Jiang, Yuchao <yuc...@email.unc.edu>
Subject: CODEX2 weird behavior

 

Dear Yuchao,

I am using your tool CODEX2 for copynumber detection in matched normal-tumor samples. I used it on three datasets which altogether contained 58 patients (116 samples). All samples are fine except one, where I have found overlapping copy number regions (see the data bellow, bold rows). Is this expected behavior of the tool or is it some kind of bug? Which copy number should I take into consideration in this case?

chr14 105864160 105895670 0.291
chr14 105844713 105860243 1.325
chr14 105860253 105863349 3.015
chr14 106680602 106875071 2.76
chr14 105837032 105845677 1.611
chr14 68861792 68925696 2.566
chr14 105863675 105864196 1.096
chr14 49643541 49655819 2.487
chr14 91338016 91462247 1.543
chr14 94045775 94227385 1.712
chr14 64024279 64141501 2.176
chr15 51958661 51958906 0.874

Thank you very much for your answer.

Yours sincerely,

Janka Puterova

Bioinformatician/PhD student at Brno University of Technology, Czechia

 

Bez virů. www.avast.com