Creating pseudo time within a time point

[fereshtehizadi]

Sorry, this is not an issue rather a question,

I am working with a mold right now without a defined list of cell cycle genes. At vegetative stage (time point 0) this organism is barely expression cell cycle genes (growing). I need to arrange cells in time point 0 by pseudo time regarding cell cycle. Is it possible only by using one time point? I have 9 time points but after time point 0 mold starts to differentiation.

could you please kindly guid me?

Thanks a lot in advance

[JohnReid]

Hi, sorry I just saw this question. No that isn't possible with DeLorean. I think Oscope is a nice method for this sort of problem. It might be worth investigating that further: https://www.nature.com/articles/nmeth.3549

[fereshtehizadi]

Thanks a lot for replying my query; Sorry, I am working an organism without any reference of cell cycle genes unlike human or mice (no Cyclebase genes for that). I ran DeLorean on my data and I saw no pattern and just human genes IDs instead of my gene IDs; So, can not somebody use DeLorean without Cyclebase genes or there is a way to use DeLorean on cell cycle pseudotime without cell cycle reference genes? Thanks a lot in advance for any help Fereshteh Izadi

…

________________________________ From: John Reid <notifications@github.com> Sent: 16 October 2018 23:18:38 To: JohnReid/DeLorean Cc: Izadi, Fereshteh; Author Subject: Re: [JohnReid/DeLorean] Creating pseudo time within a time point (#4) Hi, sorry I just saw this question. No that isn't possible with DeLorean. I think Oscope is a nice method for this sort of problem. It might be worth investigating that further: https://www.nature.com/articles/nmeth.3549 — You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub<#4 (comment)>, or mute the thread<https://github.com/notifications/unsubscribe-auth/AmWszHb6W1hcsq9kNSUr5o11lCe1ZvbEks5ulls-gaJpZM4WiySd>.

[JohnReid]

That sounds a little confusing. If you ran DeLorean on your data, you should see your gene IDs, not human IDs. Can you send a minimal example showing this?

Regarding the time points, DeLorean could fit your data in theory. In practice it uses an empirical Bayes method to set hyperparameters before fitting. This method estimates how much variation there is within time (a.k.a. capture) points and between time points. Hence DeLorean requires more than one time point in the data.

Did you look at Oscope? To me it sounds a better fit to your problem.