Clone sizes issue

[underbais]

Hi @rschenck

Nice library. A bit confused about how your package deals with clone sizes though. Looks like it blindly stacks up all clones as if they are independent (almost never the case). Below is the plot and df generated from your internal data. Clone 4 is a child of clone 3 and in the dataframe they both have the size of 300 but the plot obviously blows up the parental clone to 600. Usually clone size is derived from VAF/CF and cellularity data so having that data and clonal structure one could immediately put the evoseq df together without adjusting for 'stacking issue' in your library. So the whole tumor size at time point 8 should be just the sum of two dominating clones (clone 3 and clone 5) which is 600 according to df, but it's over 1000 in the plot. I doubt it's a feature, or is it? Thoughts?

Thanks!

[cdgatenbee]

Hello @underbais ,
Not sure what happened, but I only just now saw your Issue :( In response to your comment though, I think I can offer some clarification, but please let me know if I've misunderstood your comment. What evofreq is plotting is the total number of cells (or individuals) that carry a mutation, as opposed to individual population sizes. For example, at timepoint 8, clones 3 and 4 both have a population size of 300, and so the summed size of each population is 600. If we were plotting population sizes, then clones 3 and 4 should take up the same amount of area. However, since we are plotting the number to cells that carry a mutation (indicated by the nesting of populations), the total number of cells that have the clone 3 mutation is 600, as clone 4 carries clone 3's mutation, which is reflected in the size of clone 3 at timepoint 8. Also, the total population size at timepoint 8 is 1, 050 because we include the sizes of clones 5-8 in the plot. I hope that clarifies how we are plotting clone sizes, but if not, please let me know.

Best,
-Chandler