search

twopin / CAMP

predicting peptide-protein interactions
117 stars 30 forks source link

error in CAMP_BS.h5 modelling #33

Closed xiaoxiao349 closed 1 year ago

xiaoxiao349 commented 1 year ago

Hi when I use CAMP_BS.h5 for prediction I got this error. Could you please give some advice on it? image

denizozcivi commented 1 year ago

Hi Dear, how did you run the application? Can you please give advice to us?

twopin commented 1 year ago

Hi, what the input size you fed into the model?

xiaoxiao349 commented 1 year ago

Hi Dear, how did you run the application? Can you please give advice to us?

Hi, I started from codes in preprocess_features.py and predict_CAMP.py and used the CAMP.h5 to run the model.

xiaoxiao349 commented 1 year ago

Hi, what the input size you fed into the model?

Hi twopin thanks for replying, I input 6 protein and peptide sequence and got the 8 files after preprocessing (protein_feature_dict, peptide_feature_dict, ...). Using the function load_example I got [X_pep, X_p, X_SS_pep, X_SS_p, X_2_pep, X_2_p, X_dense_pep, X_dense_p, pep_sequence, prot_sequence, X_pep_mask, X_bs_flag]. I'm not sure what is the input size though.

twopin commented 1 year ago

Hi, I hope the discussion yesterday can solve your concerns. There are several key points may cause the difference: 1. the SS feature you generated are derived from different version of SCRATCH, after comparing we noticed that there are many differences; 2. Although we used the same PDDbench list, we used different sampling methods to generate negatives. 3. CAMP adopts UniProt sequences instead of PDB fasta sequences. 4. The Intrinsic Disorder values are different. To solve these problems, I already send you my PDDbench data (including negatives), my inference results and evaluation scripts. Hope these stuff can help you.