Rna capable

conkit/command_line/conkit_validate.py

@@ -311,6 +311,8 @@ def main():
     if args.RUN_SVM=='yes':
         logger.info(os.linesep + "Running Support Vector Machine.")
 
+        validation.calculate_features()
+
         if args.moltype=='Protein':
             p = PDBParser()
             structure = p.get_structure('structure', args.pdbfile)[0]

conkit/plot/modelvalidation.py

@@ -141,9 +141,20 @@ def __init__(self, model, prediction, sequence, map_align_exe=None, dist_bins=No
         self.prediction = prediction
         self.sequence = sequence
         self.absent_residues = self._get_absent_residues()
+        prediction_cmap = self._prepare_contactmap(self.prediction.copy())
+        predicted_dict = prediction_cmap.as_dict()
 
         self.data = pd.DataFrame()
 
+        self.data['RESNUM'] = predicted_dict.keys()
+        self.data['MISALIGNED'] = False        
+        self.data['SCORE'] = 0
+        self.data['CONTACTS'] = 0        
+        self.data['PLDDT'] = 0
+        self.data['Q_IN_ERROR'] = ''  
+
+    def calculate_features(self):
+
         model_distogram = self._prepare_distogram(self.model.copy())
         prediction_distogram = self._prepare_distogram(self.prediction.copy())
         model_cmap = self._prepare_contactmap(self.model.copy())
@@ -296,14 +307,10 @@ def _parse_data(self, predicted_dict, *metrics):
         for residue_features in zip(sorted(predicted_dict.keys()), *metrics):
             _features.append((*residue_features,))
 
-        self.data = pd.DataFrame(_features)
-        self.data.columns = ALL_VALIDATION_FEATURES
-
-        self.data['MISALIGNED'] = False        
-        self.data['SCORE'] = 0
-        self.data['CONTACTS'] = 0        
-        self.data['PLDDT'] = 0
-        self.data['Q_IN_ERROR'] = ''  
+        feature_df = pd.DataFrame(_features)
+        feature_df.columns = ALL_VALIDATION_FEATURES
+
+        self.data = self.data.merge(feature_df, how='inner', on =['RESNUM'])
 
 
 

conkit/plot/tools.py

@@ -408,7 +408,7 @@ def get_zscores(model_distogram, predicted_cmap_dict, absent_residues, *metrics)
     absent_residues: list, tuple, set
         The residues that are missing from the model
     *metrics: list
-        The mertics for which the Z-Scores will be calculated
+        The metrics for which the Z-Scores will be calculated
 
     Returns
     -------