search

PoonLab / vindels

Developing an empirical model of sequence insertion and deletion in virus genomes
1 stars 0 forks source link

Historian sanity checks #98

Closed jpalmer37 closed 3 years ago

jpalmer37 commented 3 years ago

First, I verified that all the sampled trees from BEAST analysis do in fact have the same topology, just in case.

On these identical phylogenetic trees, Historian is mostly consistent but does not give identical results. In multiple patients, it appears that the differences are caused one specific indel event (i.e. this one specific indel event is present in some trees and not in others while all other indels stay consistent). I am in the middle of reconstructing an example to further investigate these partial cases.

jpalmer37 commented 3 years ago

We opted not to continue investigating the small discrepancies (1 indel off) in Historian output when running it on identical trees (old analysis). However, I'm going to leave this open while I perform Historian sanity checks on the new results.

jpalmer37 commented 3 years ago

Completed