Hello:

I have an important doubt about the results of the comparison between both tumoral and normal samples in cn.mops. In the reference manual I saw that for this type of comparisons it is required the tuning of the referencecn.mops function in order to increase the LFC and the number of classes of the function to obtaia better recognition. In my problem, I have 5 tumoral and 5 normal samples from 5 different patients. In this sense, for 1 tumoral samples I have 1 normal samples from the same patient (cohort). The problem relies on when I run the study with cn.mops and I extract the cnvs and the cnvr and I do not obtain a class greater than CN8 in any of the cases. It is suppose that for this type of problem (tumor vs normal) I have to obtain classes greater than CN8 and I do not know if I am doing something wrong. I attach you the code that I am using for this study.

Thank you in advance for your help.

Best regards.

Daniel.

tumor_gr <- getReadCountsFromBAM(tumorBAMFiles,
                  refSeqName=paste("chr",c(1:22,"X","Y"),sep = ""),parallel = 8)

normal_gr <- getReadCountsFromBAM(normalBAMFiles,
                  refSeqName=paste("chr",c(1:22,"X","Y"),sep = ""),parallel = 8)

ref_analysis <- referencecn.mops(tumor_gr,normal_gr, I = c(0.025, 0.5, 1, 1.5, 2, 2.5, 3, 3.5, 4, 8,16,32,64), 
                                 classes = c("CN0", "CN1", "CN2", "CN3", "CN4", "CN5", "CN6", "CN7","CN8","CN16","CN32","CN64","CN128"))

resCNMOPS <- calcIntegerCopyNumbers(ref_analysis)

segplot(resCNMOPS)

cnvr <- cnvr(resCNMOPS) #look at CNV regions

cnvs <- cnvs(resCNMOPS) #look at individual CNV regions