User test module - README data/ pdb/: User input structures in pdb format ptf/: Store extracted positions of test sites centered around targeted residue type(s) The targeted residue type(s) is determined by the specified functional site of interest. - Specified conserved residue for PROSITE functional families - CYS, ARG, TRP and GLU for NOS site - HIS and SER for TRYPSIN_like enzyme detection - All residues for ATP binding residue detection numpy/: numpy arrays of extracted local protein boxes in voxel representation code/ main.py ---- The main file that the users will interact with to perform the detection tests Usage: THEANO_FLAGS=floatX=float32 python main.py module_name site_name PDB_list - module_name : Can be one of "PROSITE", "TRYPSIN_like", "NOS" or "ATP" - site_name : Use to specify functional site within the 10 PROSITE families for test If the chosen module_name is "PROSITE", please input any of the below 10 functional families 'EGF_1.10' 'IG_MHC.3' 'EF_HAND_1.1' 'RNASE_PANCREATIC.2' 'INSULIN.2' 'PROTEIN_KINASE_ST.5' 'PROTEIN_KINASE_TYR.5' 'TRYPSIN_HIS.5' 'TRYPSIN_SER.6' 'ADH_SHORT.5' If the chosen module_name is "TRYPSIN_like", "NOS" or "ATP", please repeat the module_name as site_name - PDB_list : A list of user input PDB structures. Please put the PDB files and PDB under the directory data/pdb. Example PDB list can be found in the data/pdb directory (IG_FN.list, TRYPSIN.list, NOS.list, ATP_PDB.list) Example: (example PDB structures and PDB list are provided under the data/pdb directory) THEANO_FLAGS=floatX=float32 python main.py PROSITE IG_MHC.3 IG_FN.list THEANO_FLAGS=floatX=float32 python main.py TRYPSIN_like TRYPSIN_like TRYPSIN.list THEANO_FLAGS=floatX=float32 python main.py NOS NOS NOS.list THEANO_FLAGS=floatX=float32 python main.py ATP ATP ATP_PDB.list Code files that will be called by main.py to perform individual tasks: Data processing Extract positions of test sites centered around targeted residue type, write positions in ptf files extract_all_res_defined_AA_type.py Extract positions of test sites centered around all residues, write positions in ptf files extract_all_res.py Extract local boxes around specified positions in the correpsonding pdb files cut_box_site_test_pdb.py Evaluate probability scores of each test site using trained 3DCNN models eval_ATP_3DCNN.py eval_NOS_3DCNN.py eval_PROSITE_3DCNN.py eval_TRYPSIN_3DCNN.py layers.py Record representative probabilty score for each site detection_summary.py Generate PDB files with representative probability scores of each examined site written in the B-factor field: visualize_detected_sites.py results/ prob_score/: Predicted probability scores of test sites summary/: Recorded representative probabilty score for each site pdb_detect/: PDB files with representative probability score of each examined site written in the B-factor field