Closed anushree-a closed 5 years ago
Thanks for letting me know. I wrote this example with Deeplift 0.4.0 and the structure of the Deeplift code base has changed since then. I would use the older, compatible version of Deeplift or otherwise you can update compute_importance() to respect the new structure (i.e. using deeplift.layers instead of deeplift.blobs). I will try to update the example soon.
Tried running the embedded motif interaction code for my own dataset (simulated as per the guidelines in the supplementary paper) and encountered the following error:
Traceback (most recent call last): File "embedded_motif_interaction_example.py", line 162, in
reference_gc=0)
File "build/bdist.linux-x86_64/egg/dfim/core.py", line 467, in compute_importance
AttributeError: 'module' object has no attribute 'blobs'
I am using the following software versions - Tensorflow version = 1.9.0 Keras version = 2.2.4 Deeplift version = 0.6.6.2
Please let me know how to work around it. Thank you in advance for your time!