search

saezlab / ccc_protocols

LIANA x Tensor-cell2cell Protocols
https://ccc-protocols.readthedocs.io
MIT License
1 stars 3 forks source link

Pathway issue #16

Open SimonE1220 opened 6 months ago

SimonE1220 commented 6 months ago

Hey, sorry for bothering you with more issues. I have an error with the pathway analysis could you please help me. I am running seaborn 0.11.2 on Python 3.10.11 Unbenann2t

earmingol commented 6 months ago

Unfortunately KEGG is not available for mouse, only 'GOBP' and 'Reactome' works with mouse. 'KEGG' is available only for human at the moment.

We will eventually add KEGG for mouse.

SimonE1220 commented 6 months ago

Thank you, worked now.

dbdimitrov commented 5 months ago

Hi @SimonE1220,

Great to hear that it worked. Just in case you want to use other resources, you could check this vignette for resource orthology conversion: https://decoupler-py.readthedocs.io/en/latest/notebooks/translate.html

SimonE1220 commented 5 months ago

Thanks a lot for your response. I have one more question intepreting the results. For example i have condition A and B. And than i perform the progeny analsis, it shows that Nfkb activity is upregulated wheras Jak-Stat activity is reduced. How do I know if the activity is upregulated in condition A or B, so whats the reference ? Thank you very much.

dbdimitrov commented 5 months ago

Hi @SimonE1220,

High interaction loadings (and hence pathway activities estimated on those) should be linearly associated with your sample/context loadings.

So, whichever A or B has high loadings in your factor of interest is also the condition where NFKB activity is up. The same goes for JAK-STAT - i.e. it's down in the samples with high loadings.

For the samples with low loadings, you can think of the pattern (i.e. signaling captured by Factor X) as not being strongly associated with those.

Hope this helps!

SimonE1220 commented 5 months ago

@dbdimitrov Thanks a lot that realy helps me.