search

ventolab / CellphoneDB

CellPhoneDB can be used to search for a particular ligand/receptor, or interrogate your own HUMAN single-cell transcriptomics data.
https://www.cellphonedb.org/
MIT License
333 stars 53 forks source link

pvalues.txt outputting binary results #21

Closed rs-hunter closed 2 years ago

rs-hunter commented 2 years ago

Hi,

I'm trying to generate dotplots using the pvalues.txt and means.txt files output by the statistical_analysis method. However, the output of the pvalues file appears to be a binary result rather than a p-value. I've attached one of the outputs here: pvalues.txt Is anyone else experiencing this issue? I saw a closed issue with the same concern but the poster turned out to be mistaken, so maybe it's a problem with the files I'm providing. Any suggestions would be appreciated!

Thanks, RH

luzgaral commented 2 years ago

Dear rs-hunter,

Can you provide your cellphonedb run?

Thank you

rs-hunter commented 2 years ago

Here it is: cellphonedb method statistical_analysis naive_meta.txt naive_counts.txt [ ][APP][15/03/22-10:23:30][WARNING] Latest local available version is v2.0.0, using it [ ][APP][15/03/22-10:23:30][WARNING] User selected downloaded database v2.0.0 is available, using it [ ][CORE][15/03/22-10:23:30][INFO] Initializing SqlAlchemy CellPhoneDB Core [ ][CORE][15/03/22-10:23:30][INFO] Using custom database at /Users/rahelhunter/.cpdb/releases/v2.0.0/cellphone.db [ ][APP][15/03/22-10:23:30][INFO] Launching Method cpdb_statistical_analysis_local_method_launcher [ ][APP][15/03/22-10:23:30][INFO] Launching Method _set_paths [ ][APP][15/03/22-10:23:30][WARNING] Output directory (/Users/rahelhunter/out) exist and is not empty. Result can overwrite old results [ ][APP][15/03/22-10:23:30][INFO] Launching Method _load_meta_counts [ ][APP][15/03/22-10:24:49][INFO] Launching Method _check_counts_data [ ][CORE][15/03/22-10:24:49][INFO] Launching Method cpdb_statistical_analysis_launcher [ ][CORE][15/03/22-10:24:49][INFO] Launching Method _counts_validations [ ][CORE][15/03/22-10:24:49][INFO] Launching Method get_interactions_genes_complex [ ][CORE][15/03/22-10:24:50][INFO] [Cluster Statistical Analysis] Threshold:0.1 Iterations:1000 Debug-seed:-1 Threads:4 Precision:3 [ ][CORE][15/03/22-10:24:52][INFO] Running Real Analysis [ ][CORE][15/03/22-10:24:52][INFO] Running Statistical Analysis [ ][CORE][15/03/22-11:40:43][INFO] Building Pvalues result [ ][CORE][15/03/22-11:40:43][INFO] Building results

luzgaral commented 2 years ago

According to your pvalues.txt, you only have ONE celltype and cellphone is only checking interactions within this celltype (dmm10k|dmm10k).

CellphoneDB statistical_analysis is designed to work with multiple celltypes.

rs-hunter commented 2 years ago

Got it - thank you!

raman91 commented 1 year ago

Hi, I have multiple celltypes, but still getting binary p-values :( Could you explain why it could be possible? Thank you

raman91 commented 1 year ago

pvalues.txt Here, is my file for your reference. Thanks

raman91 commented 1 year ago

Is it because of --pvalue (Pvalue threshold) option?