Bram Van de Sande

Dear, I did some research and the problem potentially resides in the dask.distributed package (see https://github.com/dask/distributed/issues/1830). The author of the dask framework, Matthew Rocklin, advises to downgrade the version of...

Dear Hicham, I tried to reproduce the problem using a fresh installation of pyscenic (version 0.8.16) in a virgin miniconda environment (on a Linux RedHat distribution running on a dual...

Dear, Good suggestion. A workaround, for now, could be to convert the motif file into regulons via a Python code snippet: ``` import pickle from pyscenic.utils import load_motifs from pyscenic.prune...

These bugs should be fixed in release 0.9.19.

Hi Gökçen, Hi Chris, You are entirely correct in stating that the whole genome rankings for all cells in an experiment for the AUCell step is done in a non-parallel...

Hi 1. Because SCENIC's first step, i.e. network inference using GENIE3/GRNBoost2, relies on tree-based methods there should be no need to transform the gene expression matrix. GENIE3 is based on...

Hi, Within the pyscenic conda environment (`activate conda pyscenic`), could you try to install and properly configure hdf5 with the following command and try pyscenic again? `conda install -c anaconda...

Dear, The problem seems to be in creation of regulons from the enriched motif table. A temporary and quick workaround could be the use the CLI command for the cisTarget...