search

saezlab / liana

LIANA: a LIgand-receptor ANalysis frAmework
https://saezlab.github.io/liana/
GNU General Public License v3.0
169 stars 30 forks source link

Error with liana_wrap() #94

Closed ChristinaDing101 closed 1 year ago

ChristinaDing101 commented 1 year ago

Hi, I encounter a problem when i am trying to run liana_wrap(). The Rstudio says as below: Running LIANA with Myeloid_Subtype as labels! Idents were converted to factor Error in h(simpleError(msg, call)) : 在为'rowMeans'函数选择方法时评估'x'参数出了错: size factors should be positive 此外: Warning message: In exec(output, ...) : 10882 genes and/or 0 cells were removed as they had no counts!

However, it is strange that when i manually aggregate some macrophage subsets into one macrophage subset, the problem seems to be gone. I wonder if it happens because some of my macrophage subset only has one to two cells?

dbdimitrov commented 1 year ago

Hi @ChristinaDing101,

I believe the small clusters should be filtered out automatically - I.e anything with less than 10 cells.

Likely there is another issue - e.g. you have negative counts of something similar.

dbdimitrov commented 1 year ago

Multiple resolutions to this error message are described here: https://github.com/saezlab/liana/issues/18

Let me know if any help, alternatively I could try to debug with a small subset of your data.

ChristinaDing101 commented 1 year ago

Hi @dbdimitrov , Thank you for your in time response. With your suggestion, I check my data, and it seems like that my data needs to be normalized first, otherwise it contains negative counts. Thanks.