search

scverse / squidpy

Spatial Single Cell Analysis in Python
https://squidpy.readthedocs.io/en/stable/
BSD 3-Clause "New" or "Revised" License
424 stars 78 forks source link

Questions about ligrec in squidgy #406

Closed HelloWorldLTY closed 2 years ago

HelloWorldLTY commented 3 years ago

Is it mainly constructed by cellphonedB? Is the result of this method same as cellphonedB result?

If I intend to find the number of significant pairs, what should I do? Thanks.

michalk8 commented 3 years ago

Hi @ChineseBest

Is it mainly constructed by cellphonedB?

We use our reimplementation of the cellphoneDB algorithm and get the receptor-ligand pairs from Omnipath (though you can also specify cellphoneDB as the only resource).

Is the result of this method same as cellphonedB result?

The best way to answer this is forr you to take a look at this thread (particularly the comment I'm linking): https://github.com/theislab/squidpy/issues/308#issuecomment-812473384 and have a look at the reproducibility notebook here. In short, the p-values correlate highly with the original cellphoneDB implementation and the means should all match.

If I intend to find the number of significant pairs, what should I do?

Please have a look at our example here: https://squidpy.readthedocs.io/en/stable/auto_examples/graph/compute_ligrec.html

Hope this answers your questions!

HelloWorldLTY commented 3 years ago

Thanks for your answer. In fact, I wonder if there are more ways to analyze cell-cell interaction without using cell type label as cluster keywords. Thanks.

giovp commented 2 years ago

Hi @ChineseBest , is this still an issue or can we close it?