Hi all, I am recently trying to use ABSOLUTE to evaluate tumor purity and ploidy. I have some questions about the plots for each model. From GenePattern (http://software.broadinstitute.org/cancer/software/genepattern/analyzing-absolute-data), I found a simple explatation of the plots. However, I am wondering whether they are valid for any ploidy.
1. Whether the alpha/2 line for the SSNV allelic fraction plot should divide the peak of clonal SSNVs density for any estimated ploidy or just diploidy?
2. Is the multiplicity density line of clonal SSNVs always expected to be close to 1 or this should depends on ploidy? As far as I know, multiplicity represents average allele-counts per cancer cell. So intuitively, I expect this value would be larger when ploidy gets larger.
3. I have WES processed by varscan2. WES data is not quite accurate for CNV calls. So I wonder whether I should depend on SCNV or SSNV scores when I choose the best model? I found it usually the case that SSNV and SCNV scores give totally different best model estimates. Furthermore, it seems that SCNV score always towards higher ploidy, while SSNV (if I am right for the above two points) always towards diploidy.
Same question also asked at CGA (http://archive.broadinstitute.org/cancer/cga/node/728). Anyone interested in this post may also check the replies there.
Thanks a lot,
Jian Y