Error in mvnX when running processSamples.R of OncoCNV

Question

[mwarrier]

Hi,

I tried to run ONCOCNV.sh (most recent version) on a set of 9 normal sample alignments to generate a baseline to call CNVs on a set of 8 tumor samples. I used an appropriate tab-delimited BED file (chr, start, end, amplicon ID, 0, and gene name), BAM alignments, and an Ensembl reference genome. All tool requirements were satisfied. However, I was getting the following error with processSamples.R:

running processSamples.R
Package 'mclust' version 5.4.5
Type 'citation("mclust")' for citing this R package in publications.
PSCBS v0.65.0 successfully loaded. See ?PSCBS for help.

Attaching package: ‘PSCBS’

The following objects are masked from ‘package:base’:

append, load
R.cache v0.13.0 (2018-01-03) successfully loaded. See ?R.cache for help.
Loading required package: lattice
Loading required package: grid
Loading required package: parallel
Error in mvnX(data = data, prior = prior) :
NA/NaN/Inf in foreign function call (arg 1)
Calls: predictClust ... eval -> eval -> mclustBIC -> mvn -> eval -> eval -> mvnX
In addition: Warning messages:
1: In chrom != "chrM" & chrom != "chrX" & chrom != "chrY" & chrom != :
longer object length is not a multiple of shorter object length
2: In chrom != "chrM" & chrom != "chrX" & chrom != "chrY" & chrom != :
longer object length is not a multiple of shorter object length
3: In chrom != "chrM" & chrom != "chrX" & chrom != "chrY" & chrom != :
longer object length is not a multiple of shorter object length
4: In rnorm(length(autoInd), 0, sdError[autoInd]) : NAs produced
Execution halted

I want to mention that the Control.stats.Processed.txt file from processControls.R contained some fields with missing values (NA) for mean and IC1-3. Replacing those values with arbitrary numbers did not seem to help. Any clues as to what could be causing this error would be much appreciated.

Thank you!

[German.M.Demidov]

I'd advice to open an issue on github. You may try to understand what was wrong, but from my experience it will take you disproportional amount of time. Yes, it seems it is an mclust error, but in general it is really stable package...

[mwarrier]

I have opened an issue on OncoCNV's github page, no responses yet.