veseshan
commented
6 years ago Hi Dario,
logR and logOR are locus level estimates from the data. They are not corrected for purity. The only correction applied is GC-correction for logR. So the sample with higher purity will show more pronounced differences in logR levels between segments and higher separation in logR. For instance, a single copy gain will show a difference in logR of log2(1.2) and log2(1.4) for 40% and 80% purity and logOR levels of log(1.4) and log(1.8) respectively. The noise within segments can be similar since that is a function of coverage. Only the copy number states (tcn and lcn) are corrected for purity.
Venkat
Hi- As per title, I'm wondering whether the logR and logOR estimates are corrected by the purity estimate. I.e. as if the reported values are those one would see if the purity were 100%.
Let's make an example... You have two tumor samples (A and B) from the same tumor tissue. Assume A and B are the same in terms of genomic profile, including copy number variations, and there are no technical artifacts in library prep & sequencing.
HOWEVER, sample A has purity 40% and sample B has purity 80%. Assuming Facets correctly estimates these purities, what would be the logR (and logOR) profiles of these two samples? Would they be the same in terms of profile and absolute values? Or sample A would have half the log values of sample B since its purity is half?
Thank you!