How do you compute the log odds of two trajectories?

[rosshandler]

Hi WOT team,

I am trying to compute the log odds in the same way you did it for the paper. It sounds trivial but I don't get it from your toy example. I can see you do some tricks for setting up colors but I don't understand how to explicitly obtain log(p/1-p) for two trajectories. Could you please elaborate on how to do it in real data?

Thanks! Ivan

[geoffschieb]

Hi Ivan,

You can compute the descendant distribution of any cell by extracting a row of the coupling matrix (for two adjacent time points this is the transport map, for longer time intervals you multiply transport maps together).

Then the probability of obtaining a fate is the fraction of descendants in that fate, so you add up the descendant mass in fate 1 and in fate 2 and divide those and take the log.

Best, Geoff

On Wed, Mar 20, 2019 at 10:28 AM rosshandler [email protected] wrote:

Hi WOT team,

I am trying to compute the log odds in the same way you did it for the paper. It sounds trivial but I don't get it from your toy example. I can see you do some tricks for setting up colors but I don't understand how to explicitly obtain log(p/1-p) for two trajectories. Could you please elaborate on how to do it in real data?

Thanks! Ivan

— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub https://github.com/broadinstitute/wot/issues/40, or mute the thread https://github.com/notifications/unsubscribe-auth/AJIoYkDxSUS5DKNvtHPkfZhh8KeLRNGkks5vYkWUgaJpZM4b_cD- .

[rosshandler]

Excellent, thanks a lot for a quick and clear answer! Can't way to try it.

Best, Ivan

[geoffschieb]

Hi Ivan,

We've just released a tutorial on WOT: https://broadinstitute.github.io/wot/tutorial/

I'm sending this to anyone who has expressed interest recently.

Best, Geoff

On Thu, Mar 21, 2019 at 9:26 AM rosshandler [email protected] wrote:

Excellent, thanks a lot for a quick and clear answer! Can't way to try it.

Best, Ivan

— You are receiving this because you commented. Reply to this email directly, view it on GitHub https://github.com/broadinstitute/wot/issues/40#issuecomment-475227696, or mute the thread https://github.com/notifications/unsubscribe-auth/AJIoYnyCcsHGWpxBxgNBrQVlIpfKHevnks5vY4iXgaJpZM4b_cD- .