[CNV] doubt with the results of cn.mops on different reference genomes

Question

[dacacis]

Hello everyone!

I am working with DNA-seq and the cn.mops R package. Firstly, i was working with BAM files aligned with the reference genome hg19 and all work fine! But now I am working with BAM aligned with the hg38 reference genome and the plots of cn.mops are a bit different. I think that this can be due to the improvements of the hg38 in comparison with hg19 but I don't know exactly if I can use cn.mops with data from hg38 genome. How shoud I interpretate them? I attach you some photos with the resultant plots.

Plot resultant from the hg19 sample

Plot resultant from the hg38 sample

Another plot from another hg38 sample

Thank you very much in advance.

Best regards!

Answer 1

[pbpanigrahi]

The manual says "The length of the initial segments is also crucial. They should be chosen such that on average 50 to 100 reads lie in one segment. The WL parameter of getReadCountsFromBAM determines this resolution.".

Since the segments you obtain in hg19 bam files vs hg38 bam files would be different, you might see different number of segments and therefore the size of your read count matrix (N*m where N: number of segments, m: number of samples) which could affect the result.

You can cross check, how many segments generated when u used hg19 and when u used hg38

[dacacis]

Thank you very much! This was the problem, when I adjust the Windows length of the plot of all the samples that come from the hg38 are plotted correctly.