nannabarnkob
commented
2 years ago Hi there.
So I just more and less randomly came across this thread, and would love to hear if you have made any progress in the matter?
I am not a true biologist, but agree that reversions are a bit of an eyesore when I inspect my trees.
It is much easier to "comprehend" something that looks like this from Simonich et al.:
(Simonich, C.A. et al. (2019) ‘Kappa chain maturation helps drive rapid development of an infant HIV-1 broadly neutralizing antibody lineage’, Nature communications, 10(1), p. 2190. doi:10.1038/s41467-019-09481-7.)
Do you have any comments on how to approach this ?
Thanks a lot for the awesome works and tools,
Nanna
Our biologist collaborators hate seeing reversions in trees, and feel that they are a clear signal of a problem in the tree.
@WSDeWitt correctly points out that under simple models, reversion mutations are no less likely than any other particular set of mutations. However, they seem like they should be less likely for models that include context sensitivity or selection.
In order to look into this issue I'd like to be able to calculate them on the ensemble of trees in an hDAG and have that be accessible through the gctree interface.