Excellent work! However, I have some problems with reproducing the results in your paper.
I follow your pipeline in README, and generate expected .bpseq, .ct, .prob for instances in PDB TS1 successfully. Then I binary the prob map with a threshold of 0.335, but I obtain metric like this: Pre-0.0402, Rec-0.1786, F1s-0.0619.
I also tried to convert the .ct file to a npy file to compute the metric, but it still kept very low values.
I am not sure whether I use the software correctly. I will really appreciate it If you can help me with the reproduction.
Thank you in advance!
Dear authors,
Excellent work! However, I have some problems with reproducing the results in your paper.
I follow your pipeline in README, and generate expected .bpseq, .ct, .prob for instances in PDB TS1 successfully. Then I binary the prob map with a threshold of 0.335, but I obtain metric like this: Pre-0.0402, Rec-0.1786, F1s-0.0619.
I also tried to convert the .ct file to a npy file to compute the metric, but it still kept very low values.
I am not sure whether I use the software correctly. I will really appreciate it If you can help me with the reproduction. Thank you in advance!