Plans to add DESeqDataSetFromTximport?

[annegreuBI]

Hi,
I was wondering if there is any plans to add a functionality similar to R's DESeqDataSetFromTximport()?
Currently salmon data cannot be used in python unless it is rounded, which is not really the correct way to do it.
Since there now is also a pytximport package I was wondering if something is in the works.
Best
Anne

[vcabeli]

Hello @annegreuBI, thank you for your interest in this package ! I think it's a very relevant request, we will take a look at https://github.com/complextissue/pytximport

Could you give some details on what you would like PyDESeq2 to implement ? I see in the R code that importing data from tximport implies two things :

• counts are rounded https://github.com/thelovelab/DESeq2/blob/30cc350d58c307a2c564c2b372ec2a50462a8097/R/AllClasses.R#L411
• if available, the average transcript factors is used in the size factor estimation https://github.com/thelovelab/DESeq2/blob/30cc350d58c307a2c564c2b372ec2a50462a8097/R/methods.R#L383

Can you think of anything else ?

I will try to make some time to work on it, or if you want to give it a go I would be happy to help with the PR.

[BorisMuzellec]

Hi @annegreuBI and @vcabeli, is this related to following issue: #305?

If so, a fairly detailed plan was designed to implement this feature. I haven't had time yet (and probably won't in the near future) to implement it myself, but I'm happy to help with any PR :).

[annegreuBI]

Hi @vcabeli and @BorisMuzellec ,
thank you for the fast reply and yes issue #305 is exactly what I was talking about.
I tried to find a mention of tximport but apparently didn't look well enough.
Looking forward to when you manage to implement it :-).

[serenalotreck]

Just want to say that I am also interested in this feature! @vcabeli do you foresee having time in the near future to work on it? I don't have a ton of familiarity with the code bases but I could take a crack at it if nobody else has the time!

EDIT: Had a chat with a colleague (@filiko) and they're going to try and tackle this!

[maltekuehl]

@serenalotreck @filiko That sounds great! As I mentioned in the original issue #305, I would also be happy to add any required metadata or variables to the output from pytximport and would be glad to help implements part of this, I just don't have the capacity at the moment to tackle it entirely. But please reach out if I can help with anything.