add '.importFeatMatrix()' and 'ignore_PolyPhen2_errors' option to core

luponzo86 · Dec 8, 2019 · a584f74 · a584f74
1 parent c43117b
commit a584f74
Show file tree

Hide file tree

Showing 2 changed files with 17 additions and 4 deletions.
diff --git a/rhapsody/features/PolyPhen2.py b/rhapsody/features/PolyPhen2.py
@@ -104,7 +104,8 @@ def _replace_strings_in_file(fname, new_fname, dict_substitutions):
 
 
 def queryPolyPhen2(filename, dump=True, prefix='pph2',
-                   fasta_file=None, fix_isoforms=False, **kwargs):
+                   fasta_file=None, fix_isoforms=False,
+                   ignore_errors=False, **kwargs):
     # original PolyPhen-2 curl command (see:
     # http://genetics.bwh.harvard.edu/pph2/dokuwiki/faq ):
     #
@@ -196,6 +197,8 @@ def queryPolyPhen2(filename, dump=True, prefix='pph2',
                 if dump:
                     outfile = f'pph2-{k}.txt'
                     _replace_strings_in_file(outfile, outfile, orig_accs)
+        elif not ignore_errors:
+            LOGGER.warn('Please check PolyPhen-2 log file')
         else:
             LOGGER.error('Please check PolyPhen-2 log file')
 

diff --git a/rhapsody/predict/core.py b/rhapsody/predict/core.py
@@ -60,7 +60,8 @@ def __init__(self, query=None, query_type='SAVs', queryPolyPhen2=True,
             'status_file_Pfam',
             'status_prefix_Uniprot',
             'status_prefix_PDB',
-            'status_prefix_Pfam']
+            'status_prefix_Pfam',
+            'ignore_PolyPhen2_errors']
         assert all([k in valid_kwargs for k in kwargs])
 
         # masked NumPy array that will contain all info abut SAVs
@@ -218,7 +219,8 @@ def queryPolyPhen2(self, query, filename='rhapsody-SAVs.txt'):
         # submit query to PolyPhen-2
         try:
             PolyPhen2_output = PolyPhen2.queryPolyPhen2(
-                SAV_file, fix_isoforms=fix_isoforms)
+                SAV_file, fix_isoforms=fix_isoforms,
+                ignore_errors=self.options.get('ignore_PolyPhen2_errors'))
         except Exception as e:
             err = (f'Unable to get a response from PolyPhen-2: {e} \n'
                    'Please click "Check Status" on the server homepage \n'
@@ -398,6 +400,14 @@ def _buildFeatMatrix(self, featset, all_features):
             feat_matrix[:, j] = array[featname]
         return feat_matrix
 
+    def importFeatMatrix(self, struct_array):
+        assert self.featMatrix is None, 'Feature matrix already set.'
+        assert self.featSet is not None, 'Feature set not set.'
+        assert self.data is not None, 'SAVs not set.'
+        assert len(struct_array) == self.numSAVs, 'Wrong length.'
+        featm = self._buildFeatMatrix(self.featSet, [struct_array])
+        self.featMatrix = featm
+
     def _calcFeatMatrix(self, refresh=False):
         assert self.data is not None, 'SAVs not set.'
         assert self.featSet is not None, 'Feature set not set.'
@@ -460,7 +470,7 @@ def exportTrainingData(self, refresh=False):
             trainData[f] = self.featMatrix[:, i]
         return trainData
 
-    def importPrecomputedFeatures(self, features_dict):
+    def importPrecomputedExtraFeatures(self, features_dict):
         assert isinstance(features_dict, dict)
         # import additional precomputed features
         default_feats = RHAPSODY_FEATS['all']