broadinstitute / wot

A software package for analyzing snapshots of developmental processes
https://broadinstitute.github.io/wot/
BSD 3-Clause "New" or "Revised" License
140 stars 34 forks source link

Interpreting WOT ancestors and trajectory probabilities #85

Closed dkeitley closed 3 years ago

dkeitley commented 4 years ago

Hi,

I'm currently following the WOT tutorials, in particular, the trajectory visualisation part in Notebook 4. I plotted the probabilities for a cell type trajectory in a different dataset and noticed that cells with high trajectory probabilities do not necessarily reside in similar UMAP positions (see below). I was wondering how this can be the case? I understand that the probabilities inferred through optimal transport are in part, influenced by Euclidean distances, so perhaps naively, I was expecting the trajectory to move relatively smoothly across the UMAP. I computed mean Euclidean distances with the cell type of interest (circled below) and the cluster highlighted on the right has a much larger value than the neighbouring cell types on the left-hand side. What might be causing the cells on the right to be marked as ancestors in this case?

Many thanks,

Dan