search

mskcc / facets

Algorithm to implement Fraction and Copy number Estimate from Tumor/normal Sequencing.
135 stars 65 forks source link

Ploidy and purity estimates #143

Open kw10 opened 4 years ago

kw10 commented 4 years ago

Following the example usage in the vignette, I retrieve the purity and ploidy estimates:

fit = emcncf(oo)
fit$purity
fit$ploidy

For some samples I get "75%" preceding the value. For example:

fit$purity
      75% 
0.3225428
fit$ploidy
      75% 
2.553525

For most samples, I only get the value. Can you tell me what the "75%" means?

Thanks

HenrikBengtsson commented 2 years ago

Can you tell me what the "75%" means?

It has nothing to do with tumor purity or ploidy. Instead, it stems from the fact that these parameters are estimated using the 75% quantile of some data, e.g.

> quantile(1:10, probs = 0.75)
 75% 
7.75

You can safely ignore the "75%" names attribute by, for instance, doing:

fit$purity <- unname(fit$purity)
fit$ploidy <- unname(fit$ploidy)

This is ideally something emcncf() should do internally to avoid this confusion.