search

FunGeST / Palimpsest

An R package for studying mutational signatures and structural variant signatures along clonal evolution in cancer.
68 stars 19 forks source link

Cancer cell fraction > 1.0 #2

Closed wschen closed 6 years ago

wschen commented 6 years ago

I have generated clonality plots using the cnaCCF_plots() function and noticed that the cancer cell fraction histogram tends to show that values lie within the range of [0, 1.5]. This is also true for the plot on page 14 of the vignette included in the package. It seems strange for CCF to be >1.0 since the fraction of all tumor cells harboring a mutation can't be >100%, so I am wondering whether this is a result of a bug in the code or whether there is a different way to interpret CCF.

Thanks!

screen shot 2018-06-15 at 3 04 17 pm
northerngasc commented 6 years ago

Thank you for your message. Indeed a CCF >1 has no biological meaning. The CCF reported by Palimpsest is directly the result of the VAF corrected for copy-number and tumor cell content. So due to random sampling of sequenced reads the clonal mutations will typically appear as a peak centered on 1, corresponding to the VAF peak around 0.5. A CCF >1 should thus be interpreted as =1. I hope it helps! Eric