DPT coordinates for projected data

[SoloveyMaria]

Is there any option that I can get the DPT coordinates for the data that I project atop some other dataset with the dm_predict function?

[flying-sheep]

dm_predict returns the diffusion map coordinates.

What do you mean by “DPT coordinates”? DPT is a pseudo time measure, not a coordinate.

Do you want to determine the DPT from your projected cells to your original cells? That sounds useful!

[SoloveyMaria]

Hi!

On 5. Feb 2019, at 10:23, Philipp A. ***@***.***> wrote: dm_predict returns the diffusion map coordinates. What do you mean by “DPT coordinates”? DPT is a pseudo time measure, not a coordinate.

Sorry for the misleading term! I just mean DPT

Do you want to determine the DPT from your projected cells to your original cells? That sounds useful!

Yes exactly, so by now I can calculate DPT of the reference data set (with "dpt" function from the dpt package), and separately project my dataset of interest atop the reference dataset (with "dm_predict" function), and it would be nice if after I have done the projection, I could see what would be the DPT of my dataset of interest (as if it were a part of the reference data set). For instance, if I see that the dataset of interest is projected atop one of the cell populations in the reference data set, but slightly shifted along the pseudo time trajectory (e.g. because the cells are slightly older), then it would be nice if I could make a box plot of DPTs for this population in the data of interest, as well as each population of the reference dataset, so the the reader sees the shift. (The reason for this is that if I do a scatterplot, the populations overlap and it is not quite clear, how big the shift actually is)

…

— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub <#16 (comment)>, or mute the thread <https://github.com/notifications/unsubscribe-auth/ApAsjADEBRbnYAsRnG64NfDLT6Glh7taks5vKU2jgaJpZM4ahgmQ>.