I am currently working with Level 1-3 TCGA data. I'm trying to redo the Copy Number Variation analysis that they did, getting a report of copy number losses and gains, as well as loss of heterozygosity reports. Every program I have tried reports a heatmap or other visualizations for these.
What I'd like to see is a program that spits out how many individuals have gains/losses on each chromosome arm, where I can set a threshold. Another thing that would be nice is one that can identify copy-number neutral LOH (where the allele difference plot has a line at +1 and -1, but nothing at 0...in other words, no heterozygotes in a region, but no change in copy number).
Anyone know of a program that does this for me, or am I stuck 1) going through samples one at a time or 2) creating code myself to quantify this?
Thanks in advance,
Gaius
2 answers
What format is the TCGA data you're using? If you can get their copy number calls in SEG format, the rest is not too difficult to script, although you are probably going to have to do some scripting. In R, the Bioconductor package GenomicRanges makes it relatively easy to combine, compare, and filter the datasets manually.
If you need to detect LOH regions from scratch, using each sample's VCF and either segmented copy ratios (SEG) or raw array CGH data, there are a few packages to try, e.g.: ASCAT, PSCBS, TitanCNV.
pennCNV
Log in to answer this question.