How to generate the pdbpath in input file?

[Admire7494]

I found that the output will be different as the pdbpath has changed.
Most important, as the pdbpath is same(e.g. seq1.pdb),the output will be similar although the sequences are different.

[Admire7494]

I have found the same situation in the colab for single enzyme-substrate prediction.
When using the different Uniprot ID(enzyme 1 and enzyme 2), it leads to different output for the same sequence and substrate.
When using the same Uniprot ID for different enzyme and the same substrate, it always leads to extremely similar output.