callLOH output file NA values
1
0
Entering edit mode
twtoal • 0
@twtoal-15473
Last seen 6 months ago

I updated to a newer version of PureCN and ran into a problem in my post-processing of PureCN output from callLOH. The problem was caused by NA values in the "M" column. My previous PureCN version never had NA values in the column. The PureCN doc file doesn't mention NA in its description of the "M" column. Is there a problem, or is NA expected and just means PureCN is unable to estimate M?

I'd like to suggest whenever possible, the doc files should be updated to indicate changes such as this, and furthermore, since they may have an effect on the user's downstream processing, they should be mentioned in the release notes. Thanks.

PureCN NA values • 193 views
0
Entering edit mode
@markusriester-9875
Last seen 28 days ago

Yes, the callLOH function changed and I actually mention this in the NEWS file ( http://bioconductor.org/packages/release/bioc/news/PureCN/NEWS ).

LOH calling requires heterozygous SNPs, so NA means no SNPs. The change of defaults was necessary because some users were not aware that segments without SNPs were missing from this output.

0
Entering edit mode

Ok, thanks. It didn't occur to me that that change might introduce NAs in result. Since I don't call callLOH() directly, but use PureCN.R to do it for me, I'm not real tuned in to those function names. What might be helpful is to mention output filenames in the release notes, e.g.:

callLOH (PureCN.R output file *_loh.csv) now reports all segments, even if there are no informative SNPs (indicated by NA in some output columns) since some users were not aware that segments are missing from this output. Use keep.no.snp.segments = FALSE to restore old behaviour.

0
Entering edit mode

It looks like the keep.no.snp.segments = FALSE argument to callLOH() has no equivalent option argument in PureCN.R ?