search

VanLoo-lab / ascat

ASCAT R package
https://www.mdanderson.org/research/departments-labs-institutes/labs/van-loo-laboratory/resources.html#ASCAT
162 stars 85 forks source link

`ascat.runAscat` "vector size cannot be NA/NaN" and/or "subscript out of bounds" #173

Closed MikeWLloyd closed 4 months ago

MikeWLloyd commented 5 months ago

I am working on CNV array data derived from an Illumina HumanCytoSNP-12 v2.1 array run. I am trying to understand why one sample is failing in the ascat.runAscat step.

I do not know the gender of the individual sample. So I have run ascat.runAscat as both XY and XX.

When run as XY, I have the following error in the ascat.runAscat command:

Error in sample.int(length(x), size, replace, prob) : 
  vector size cannot be NA/NaN

When run as XX, I have the following error in the ascat.runAscat command:

Error in m[i, 2] : subscript out of bounds

I am enclosing output from the run, which may help. Can you help me to understand why this sample is failing or things to further look at to narrow in on the issue for this sample?

run_output.zip

zhangzhhcb commented 4 months ago

Hi, the BAF plots in your folder are strange. It seems that very limited number of heterozygous SNPs across all chromosomes, or all chromosomes have LOH, which is impossible. Please check your raw data and you expect to get BAF like figure 3b in this paper: https://www.nature.com/articles/s41592-020-01013-2.

MikeWLloyd commented 4 months ago

Thank you for the reply. Given your answer, and the number of samples I have seen that do not have this problem, I suspect this sample was problematic either in prep or on the array.