tf_result_explanations - AttributeError: module 'networkx' has no attribute 'read_gpickle'

[sngholee]

Hello,

I'm attempting to run tf_results_explanations. I successfully ran the previous two steps

tf_result_explanations -p 0.01 -d High_ATAC_0_peaks_vs_High_ATAC_24_peaks_differential_md_scores.txt -o High_0vs24_explanations

But I get this error:

Traceback (most recent call last): File "/public/home/slee393/miniconda3/bin/tf_result_explanations", line 8, in <module> sys.exit(tf_result_explanations()) ^^^^^^^^^^^^^^^^^^^^^^^^ File "/public/home/slee393/miniconda3/lib/python3.11/site-packages/DAStk/__init__.py", line 21, in tf_result_explanations e.main() ^^^^^^^^ File "/public/home/slee393/miniconda3/lib/python3.11/site-packages/DAStk/tf_result_explanations.py", line 36, in main G = nx.read_gpickle(KNOWLEDGE_GRAPH) ^^^^^^^^^^^^^^^ AttributeError: module 'networkx' has no attribute 'read_gpickle'

Any help would be appreciated. Is this a Python version issue?

[ignaciot]

It looks like the networkX library changed the methods to read/write pickled graphs in version 3.0. I'll be pushing an update to support this, but in the meantime you can test downgrading your networkx version to any one 2.9 or lower.