Repository cleanup

.gitignore

@@ -125,3 +125,6 @@ dmypy.json
 
 #Unfinished postgres scripts
 /src/data/postgres/
+
+# Dependency directories
+src/web/node_modules/

src/api/Makefile

@@ -0,0 +1,6 @@
+data:
+	python data/make_dataset.py
+	python features/build_features
+
+train:
+	python models/train_model.py

src/api/requirements.txt

@@ -0,0 +1 @@
+flask==1.1.1

data/similarity.py → src/api/similarity.py

@@ -1,124 +1,124 @@
-""" class definitions for preprocessing target data
-"""
-import os
-
-import numpy as np
-from rdkit import Chem
-from rdkit import DataStructs
-from rdkit.Chem import PandasTools, rdFingerprintGenerator
-import pandas as pd
-from rdkit.Chem.Fingerprints import FingerprintMols
-
-
-def update_csv(df):
-    path_to_download = os.path.normpath(os.getcwd() + os.sep + '/raw/')
-    df.to_csv(r'' + path_to_download + '/{}'.format(str(df.name)),
-              sep=',', encoding='utf-8', index=True, index_label='id')
-
-class ChemblDataBatch(object):
-    """  base class that reads the .csv files on /raw/ folder using os.paths
-        Uses similiarity threshold = 1 for now
-    """
-
-    def __init__(self, data1=None, data2=None, threshold=1, **kwargs):
-        self.sim_threshold = threshold
-        path_to_data = os.path.normpath(os.getcwd() + os.sep + 'raw/')
-        csv_targets = os.listdir(path_to_data)
-        csv_targets = filter(lambda x: x.startswith("CHEMBL"), csv_targets)
-        df_returned = []
-        for element in csv_targets:
-            df = pd.read_csv('' + path_to_data + '/{}'.format(element), sep=',', encoding='utf-8', index_col='id')
-            # Droping null values from canonical_smiles
-            df = df.dropna(subset=['canonical_smiles'])
-            df.name = '{}'.format(str(element))
-            df_returned.append(df)
-            print('Read: {}'.format(element))
-
-        self.number_of_elements = (len(df_returned))
-        # TODO what if theres more than 2 here:
-        self.data0 = df_returned[0]
-        self.data1 = df_returned[1]
-        self.sim_list = []
-
-
-    def similarity_list(self, **kwargs):
-        """ used to reset our internal state so that iteration
-          starts again from the beginning
-        """
-        # Get dataframe names
-        dataframes_read = []
-        for x in range((int(self.number_of_elements))):
-            dataframes_read.append('data{}'.format(x))
-        # Iterate over multiple files
-        data_to_compare = []
-        c_smiles = []
-        for a in dataframes_read:
-            data_to_compare.append(getattr(self, a))
-        for i, a in enumerate(data_to_compare):
-            for ds in data_to_compare[i]['canonical_smiles']:
-                cs = Chem.CanonSmiles(ds)
-                c_smiles.append(cs)
-            molecules = ([Chem.MolFromSmiles(x) for x in data_to_compare[i]['canonical_smiles']])
-            fps = [FingerprintMols.FingerprintMol(x) for x in molecules]
-
-            # the list for the dataframe
-            qu, ta, sim = [], [], []
-
-        # compare all fp pairwise without duplicates on the same csv
-            for n in range(len(fps)):
-                s = DataStructs.BulkTanimotoSimilarity(fps[n],
-                                                       fps[n + 1:])  # +1 compare with the next to the last fp
-                # collect the SMILES and values
-                for m in range(len(s)):
-                    if s[m] == self.sim_threshold:
-                        qu.append(c_smiles[n])
-                        ta.append(c_smiles[n + 1:][m])
-                        sim.append(s[m])
-
-            d = {'query_{}'.format(data_to_compare[i].name): qu,
-                 'target_{}'.format(data_to_compare[i].name): ta,
-                 'Similarity': sim}
-            d = pd.DataFrame(data=d)
-            d.name = 'sim_{}'.format(data_to_compare[i].name)
-            self.sim_list.append(d)
-            update_csv(d)
-
-            # TODO make an abstract way of generating similarity for both files
-            c_smiles1 = []
-            c_smiles2 = []
-            for ds in data_to_compare[0]['canonical_smiles']:
-                cs = Chem.CanonSmiles(ds)
-                c_smiles1.append(cs)
-            for ds in data_to_compare[1]['canonical_smiles']:
-                cs = Chem.CanonSmiles(ds)
-                c_smiles2.append(cs)
-            molecules1 = ([Chem.MolFromSmiles(x) for x in data_to_compare[0]['canonical_smiles']])
-            molecules2 = ([Chem.MolFromSmiles(x) for x in data_to_compare[1]['canonical_smiles']])
-            fps1 = [FingerprintMols.FingerprintMol(x) for x in molecules1]
-            fps2 = [FingerprintMols.FingerprintMol(x) for x in molecules2]
-            # the list for the dataframe
-            qu, ta, sim = [], [], []
-
-            # compare all fp pairwise without duplicates on the same csv
-            for n in range(len(fps1)):
-                s = DataStructs.BulkTanimotoSimilarity(fps1[n],
-                                                       fps2[:])
-                # collect the SMILES and values
-                for m in range(len(s)):
-                    if s[m] == self.sim_threshold:
-                        qu.append(c_smiles1[n])
-                        ta.append(c_smiles1[n])
-                        sim.append(s[m])
-            name_file_both_sim = 'both_files.csv'
-            d = {'query_{}'.format(name_file_both_sim): qu,
-                 'target_{}'.format(name_file_both_sim): ta,
-                 'Similarity': sim}
-            d = pd.DataFrame(data=d)
-            d.name = 'sim_{}'.format(name_file_both_sim)
-            self.sim_list.append(d)
-            update_csv(d)
-
-
-obj = ChemblDataBatch()
-ChemblDataBatch.similarity_list(obj)
+""" class definitions for preprocessing target data
+"""
+import os
+
+import numpy as np
+from rdkit import Chem
+from rdkit import DataStructs
+from rdkit.Chem import PandasTools, rdFingerprintGenerator
+import pandas as pd
+from rdkit.Chem.Fingerprints import FingerprintMols
+
+
+def update_csv(df):
+    path_to_download = os.path.normpath(os.getcwd() + os.sep + '/raw/')
+    df.to_csv(r'' + path_to_download + '/{}'.format(str(df.name)),
+              sep=',', encoding='utf-8', index=True, index_label='id')
+
+class ChemblDataBatch(object):
+    """  base class that reads the .csv files on /raw/ folder using os.paths
+        Uses similiarity threshold = 1 for now
+    """
+
+    def __init__(self, data1=None, data2=None, threshold=1, **kwargs):
+        self.sim_threshold = threshold
+        path_to_data = os.path.normpath(os.getcwd() + os.sep + 'raw/')
+        csv_targets = os.listdir(path_to_data)
+        csv_targets = filter(lambda x: x.startswith("CHEMBL"), csv_targets)
+        df_returned = []
+        for element in csv_targets:
+            df = pd.read_csv('' + path_to_data + '/{}'.format(element), sep=',', encoding='utf-8', index_col='id')
+            # Droping null values from canonical_smiles
+            df = df.dropna(subset=['canonical_smiles'])
+            df.name = '{}'.format(str(element))
+            df_returned.append(df)
+            print('Read: {}'.format(element))
+
+        self.number_of_elements = (len(df_returned))
+        # TODO what if theres more than 2 here:
+        self.data0 = df_returned[0]
+        self.data1 = df_returned[1]
+        self.sim_list = []
+
+
+    def similarity_list(self, **kwargs):
+        """ used to reset our internal state so that iteration
+          starts again from the beginning
+        """
+        # Get dataframe names
+        dataframes_read = []
+        for x in range((int(self.number_of_elements))):
+            dataframes_read.append('data{}'.format(x))
+        # Iterate over multiple files
+        data_to_compare = []
+        c_smiles = []
+        for a in dataframes_read:
+            data_to_compare.append(getattr(self, a))
+        for i, a in enumerate(data_to_compare):
+            for ds in data_to_compare[i]['canonical_smiles']:
+                cs = Chem.CanonSmiles(ds)
+                c_smiles.append(cs)
+            molecules = ([Chem.MolFromSmiles(x) for x in data_to_compare[i]['canonical_smiles']])
+            fps = [FingerprintMols.FingerprintMol(x) for x in molecules]
+
+            # the list for the dataframe
+            qu, ta, sim = [], [], []
+
+        # compare all fp pairwise without duplicates on the same csv
+            for n in range(len(fps)):
+                s = DataStructs.BulkTanimotoSimilarity(fps[n],
+                                                       fps[n + 1:])  # +1 compare with the next to the last fp
+                # collect the SMILES and values
+                for m in range(len(s)):
+                    if s[m] == self.sim_threshold:
+                        qu.append(c_smiles[n])
+                        ta.append(c_smiles[n + 1:][m])
+                        sim.append(s[m])
+
+            d = {'query_{}'.format(data_to_compare[i].name): qu,
+                 'target_{}'.format(data_to_compare[i].name): ta,
+                 'Similarity': sim}
+            d = pd.DataFrame(data=d)
+            d.name = 'sim_{}'.format(data_to_compare[i].name)
+            self.sim_list.append(d)
+            update_csv(d)
+
+            # TODO make an abstract way of generating similarity for both files
+            c_smiles1 = []
+            c_smiles2 = []
+            for ds in data_to_compare[0]['canonical_smiles']:
+                cs = Chem.CanonSmiles(ds)
+                c_smiles1.append(cs)
+            for ds in data_to_compare[1]['canonical_smiles']:
+                cs = Chem.CanonSmiles(ds)
+                c_smiles2.append(cs)
+            molecules1 = ([Chem.MolFromSmiles(x) for x in data_to_compare[0]['canonical_smiles']])
+            molecules2 = ([Chem.MolFromSmiles(x) for x in data_to_compare[1]['canonical_smiles']])
+            fps1 = [FingerprintMols.FingerprintMol(x) for x in molecules1]
+            fps2 = [FingerprintMols.FingerprintMol(x) for x in molecules2]
+            # the list for the dataframe
+            qu, ta, sim = [], [], []
+
+            # compare all fp pairwise without duplicates on the same csv
+            for n in range(len(fps1)):
+                s = DataStructs.BulkTanimotoSimilarity(fps1[n],
+                                                       fps2[:])
+                # collect the SMILES and values
+                for m in range(len(s)):
+                    if s[m] == self.sim_threshold:
+                        qu.append(c_smiles1[n])
+                        ta.append(c_smiles1[n])
+                        sim.append(s[m])
+            name_file_both_sim = 'both_files.csv'
+            d = {'query_{}'.format(name_file_both_sim): qu,
+                 'target_{}'.format(name_file_both_sim): ta,
+                 'Similarity': sim}
+            d = pd.DataFrame(data=d)
+            d.name = 'sim_{}'.format(name_file_both_sim)
+            self.sim_list.append(d)
+            update_csv(d)
+
+
+obj = ChemblDataBatch()
+ChemblDataBatch.similarity_list(obj)
 print((obj.sim_list[0].name))