Confusing definition of CCF_CNV?

[lincj1994]

Hi. @luisgls @hanasusak
I want to estimate CCF using the CCF function but I found it required a input of CCF_CNV, which is defined as:
cancer cell fraction of somatic SNV in region with reported SNV.
I was truly confused since the CCF of each SNV was what I want from the CCF function. How to get the CCF_CNV or is it the same as purity? Why do all CCF_CNV equal 1 when ploidy is 2? Any correlation to nMajor or nMinor?
Thanks.

[luisgls]

Hi @lincj1994 !

So thanks for using our tool!, In our method we consider that a somatic point mutation (mut/ref) inside a CNV region can have different CCFs. So for example a deletion (copy number = 1) can be present in 70% of the cells in your bulk (if you use CNVs to estimate purity for example, this will be equivalent to purity for the somatic SNVs in that deletion), therefore your CNV or ploidy = 1 and your CCF = 0.7. A diploid region in the tumor population will then be ploidy 2 and CCF_CNV = 1, because we assume normal cells are also diploid. In summary, if you have B-allele frequency data you can then determine a priori the ploidy and the CCF of that ploidy number before calculating the CCF for the somatic point mutation. We never tried to correlate with nMajor and nMinor, but I presume if the nMajor is 2 and nMinor is 1, it refers to a triploid loci (ploidy 3) and possibly a CCF_CNV associated to the purity.