Slingshot native results vs dyn

[royfrancis]

Here is slingshot result run manually.

Here is slingshot run using dyn on the same data.

I think they look different and I am not sure what to make of it. Any thoughts?

[rcannood]

This is because a previous version (or maybe still the master version of dyno?) did not return the segment positioning of Slingshot, only the way that milestones are connected. What you are seeing is a projecton of the Slingshot trajectory onto the provided dimred.

Which version of dynmethods and dynplot are you using?

[rcannood]

Update: nevermind, not even the current version is returning this info (https://github.com/dynverse/ti_slingshot/blob/master/package/R/ti_slingshot.R#L193). I'll fix this asap.

However, please note that this only affects the plot_dimred visualisation ; other downstream analyses of the trajectory are unaffected by this.