issue running cistarget using human dataset #87
Labels
question
Further information is requested
Comments
|
Hi @sid5427 From the error I suspect that On your question wether it is possible to run SCENIC+ with a couple of the partial result. This is possible, you can generate the menr dictionary like this (in your case): import dill
CTX_topics_otsu_All = dill.load(open('results/motifs/CTX_topics_otsu_All.pkl', 'rb'))
DEM_topics_otsu_All = dill.load(open('results/motifs/DEM_topics_otsu_All.pkl', 'rb'))
CTX_topics_top_3_All = dill.load(open('results/motifs/CTX_topics_top_3_All.pkl', 'rb'))
DEM_topics_top_3_All = dill.load(open('results/motifs/DEM_topics_top_3_All.pkl', 'rb'))
menr['CTX_topics_otsu_All'] = CTX_topics_otsu_All
menr['DEM_topics_otsu_All'] = DEM_topics_otsu_All
menr['CTX_topics_top_3_All'] = CTX_topics_top_3_All
menr['DEM_topics_top_3_All'] = DEM_topics_top_3_AllBest, Seppe |
|
Hi Seppe, That's the weird part - when I run the code section for finding DARs in markers_dict I get this error - However if I run I went ahead and printed the output of |
|
Hi @sid5427 Yes indeed, it's these empty dataframes in You should remove this prior to running: for DAR in markers_dict.keys():
regions = markers_dict[DAR].index[markers_dict[DAR].index.str.startswith('chr')] #only keep regions on known chromosomes
#print(regions)
region_sets['DARs'][DAR] = pr.PyRanges(region_names_to_coordinates(regions))You can also do it like this for DAR in markers_dict.keys():
regions = markers_dict[DAR].index[markers_dict[DAR].index.str.startswith('chr')] #only keep regions on known chromosomes
if len(regions) > 0:
region_sets['DARs'][DAR] = pr.PyRanges(region_names_to_coordinates(regions))The reason that these dataframes are empty is because no regions passed the thresholds (i.e. log 2 Fold Change of 1.5 and adjusted p value < 0.05, by default). You can also change these thresholds in Best, Seppe |
|
Hi Seppe, Thanks for the solution - I'll incorporate that into my run. I had tried this to remove the three troublesome clusters - This did work, and it generated a scenicplus object with some of the downstream figures. However I get an error later for this part - and this is the error - Is this related to my ad-hoc solution? Will using the code snippet you provided solve this error downstream? Appreciate the help! |
|
Hi @sid5427 This is a known "bug" that is caused by the fact that you have an annotation ( However the fact that you're at this step means that SCENIC+ has indeed worked successfully. You can skip this optional step for now by setting Best, Seppe |
|
Hi, you can use this 6b4bdad function instead. It does not require generating pseudobulks beforehand. Best, Seppe |
|
Same problem. I don't have menr.pkl and DEM_*_topics.pkl after running run_pycistarget. |
|
Did you have any error messages after running Best, Seppe |
|
Thanks @SeppeDeWinter using a single core, it worked! |
|
You did not see any error messages using a single core? Best, Seppe |
|
Oddly, it worked without errors and gave me in output of all the files using a single core. |
Similar problem, my markers_dict is empty, which may be the cause of the death of core while running run_pycistarget. And it did not create CTX_topics_otsu_All.pkl as well as other pkl files. Instead, I only have CTX_topics_otsu_All files, should I combine all the html files and turn into a pkl and then run the above code? |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Hi Seppe and other devs - Happy new year!
I am unfortunately facing another issue while running scenic with our human 10x multiome data.
When I run pycistarget it's throwing an error - "ValueError: A gene signature must have at least one gene."
Python version - 3.8.13
Scenic version - (not sure .. Updated to latest version on december 30th - version returns - AttributeError: module 'scenicplus' has no attribute 'version' <- might want to check this as well.)
This is how I am setting up pycistarget to run ...
output log -
I looked at other error reports- namely - #60
and tried the same command with "save partial = TRUE" and "run without promoters - TRUE"
I get the same error - error log
Would it be possible to create the scenic object and run the scenic+ function with a couple of the partial result pickle files e.g. CTX_topics_otsu_All.pkl instead of menr.pkl?
Thanks!
Sid.
The text was updated successfully, but these errors were encountered: