UnpicklingError: could not find MARK

[NickEdmunds]

Describe the bug
A clear and concise description of what the bug is. UnpicklingError: could not find MARK occurs when attempting to run iScore.predict using the trained model.

Environment:

• OS system: Ubuntu
• Version: 20.04
• Branch commit ID:
• Inputs:

To Reproduce
Steps/commands to reproduce the behaviour:

1. created pdb and pssm folders under iScore/test directory
2. typed either iScore.predict or mpiexec -n iScore.predict.mpi within the test directory (both gave the same error)

Expected Results
A. Expected the creating of iScore.predict.txt file

Actual Results or Error Info
nick@nick-HP-ZBook-15-G5:~/iScore/test$ iScore.predict
/home/nick/.local/lib/python3.8/site-packages/skcuda/cublas.py:284: UserWarning: creating CUBLAS context to get version number
warnings.warn('creating CUBLAS context to get version number')
Reusing graphs in ./graph/
Reusing kernels in ./kernel/
Traceback (most recent call last):
File "/home/nick/.local/bin/iScore.predict", line 7, in
exec(compile(f.read(), file, 'exec'))
File "/home/nick/iScore/bin/iScore.predict", line 68, in
iscore_svm(load_model=None,package_name=args.archive,maxlen=args.maxlen,testID=args.ground_truth)
File "/home/nick/iScore/iScore/graphrank/rank.py", line 400, in iscore_svm
testdata = DataSet(trainID,Kfile,maxlen,testID=testID)
File "/home/nick/iScore/iScore/graphrank/rank.py", line 39, in init
self.get_K_matrix()
File "/home/nick/iScore/iScore/graphrank/rank.py", line 104, in get_K_matrix
K.update(pickle.load(open(f,'rb')))
_pickle.UnpicklingError: could not find MARK

Additional Context
Example of pdb and pssm files attached.
1084_TS029_1o.zip

[NicoRenaud]

Thanks for filling in the issue we will look into it