[RELEASE] 1.0.7

secat/main.py

@@ -354,7 +354,7 @@ def export(infile, maximum_interaction_qvalue):
 @click.option('--id', required=False, type=str, help='Plot specific UniProt bait_id (Q10000) or interaction_id (Q10000_P10000)')
 @click.option('--max_qvalue', default=0.01, show_default=True, type=float, help='Maximum q-value to plot baits or interactions.')
 @click.option('--min_abs_log2fx', default=1.0, show_default=True, type=float, help='Minimum absolute log2 fold-change for integrated nodes.')
-@click.option('--mode', default='quantitative', show_default=True, type=click.Choice(['quantitative', 'detection_integrated', 'detection_separate']), help='Select mode to order interaction plots by. Note: detection_separate will also report decoys')
+@click.option('--mode', default='quantitative', show_default=True, type=click.Choice(['quantitative', 'detection']), help='Select mode to order interaction plots by.')
 @click.option('--combined/--no-combined', default=False, show_default=True, help='Select interactions and baits according to combined q-values.')
 @click.option('--peptide_rank', default=6, show_default=True, type=int, help='Number of most intense peptides to plot.')
 def plot(infile, level, id, max_qvalue, min_abs_log2fx, mode, combined, peptide_rank):

secat/plot.py

@@ -167,12 +167,10 @@ def read_interactions(self):
 
         if check_sqlite_table(con, 'EDGE') and self.mode == 'quantitative':
             df = pd.read_sql('SELECT DISTINCT bait_id || "_" || prey_id AS interaction_id, 0 as decoy FROM %s WHERE pvalue_adjusted < %s AND abs_log2fx > %s ORDER BY pvalue ASC;' % (table, self.max_qvalue, self.min_abs_log2fx), con)
-        elif self.mode == 'detection_integrated':
+        elif self.mode == 'detection':
             df = pd.read_sql('SELECT DISTINCT bait_id || "_" || prey_id AS interaction_id, decoy FROM FEATURE_SCORED_COMBINED WHERE qvalue < %s GROUP BY bait_id, prey_id ORDER BY qvalue ASC;' % (self.max_qvalue), con)
-        elif self.mode == 'detection_separate':
-            df = pd.read_sql('SELECT DISTINCT bait_id || "_" || prey_id AS interaction_id, decoy FROM FEATURE_SCORED WHERE qvalue < %s ORDER BY qvalue ASC;' % (self.max_qvalue), con)
         else:
-            sys.exit("Mode for interaction plotting not supported.")
+            sys.exit("Error: Mode for interaction plotting not supported.")
 
         con.close()
 
@@ -183,9 +181,9 @@ def read_interactions_dmeta(self):
 
         if check_sqlite_table(con, 'COMPLEX_QM') and (self.mode == 'quantitative'):
             df = pd.read_sql('SELECT FEATURE_SCORED_COMBINED.bait_id AS bait_id, FEATURE_SCORED_COMBINED.prey_id AS prey_id, FEATURE_SCORED_COMBINED.bait_id || "_" || FEATURE_SCORED_COMBINED.prey_id AS interaction_id, BAIT_META.protein_name AS bait_name, PREY_META.protein_name AS prey_name, min(FEATURE_SCORED_COMBINED.pvalue) AS pvalue, min(FEATURE_SCORED_COMBINED.qvalue) AS qvalue FROM FEATURE_SCORED_COMBINED INNER JOIN (SELECT * FROM PROTEIN) AS BAIT_META ON FEATURE_SCORED_COMBINED.bait_id = BAIT_META.protein_id INNER JOIN (SELECT * FROM PROTEIN) AS PREY_META ON FEATURE_SCORED_COMBINED.prey_id = PREY_META.protein_id INNER JOIN (SELECT DISTINCT bait_id, prey_id FROM COMPLEX_QM) AS COMPLEX_QM ON FEATURE_SCORED_COMBINED.bait_id = COMPLEX_QM.bait_id AND FEATURE_SCORED_COMBINED.prey_id = COMPLEX_QM.prey_id GROUP BY FEATURE_SCORED_COMBINED.bait_id, FEATURE_SCORED_COMBINED.prey_id;', con)
-        elif self.mode == 'detection_integrated':
+        elif self.mode == 'detection':
             df = pd.read_sql('SELECT FEATURE_SCORED_COMBINED.bait_id AS bait_id, FEATURE_SCORED_COMBINED.prey_id AS prey_id, FEATURE_SCORED_COMBINED.bait_id || "_" || FEATURE_SCORED_COMBINED.prey_id AS interaction_id, BAIT_META.protein_name AS bait_name, PREY_META.protein_name AS prey_name, min(FEATURE_SCORED_COMBINED.pvalue) AS pvalue, min(FEATURE_SCORED_COMBINED.qvalue) AS qvalue FROM FEATURE_SCORED_COMBINED INNER JOIN (SELECT * FROM PROTEIN) AS BAIT_META ON FEATURE_SCORED_COMBINED.bait_id = BAIT_META.protein_id INNER JOIN (SELECT * FROM PROTEIN) AS PREY_META ON FEATURE_SCORED_COMBINED.prey_id = PREY_META.protein_id GROUP BY FEATURE_SCORED_COMBINED.bait_id, FEATURE_SCORED_COMBINED.prey_id;', con)
-        elif self.mode == 'detection_separate':
+        else:
             df = None
 
         con.close()
@@ -223,6 +221,9 @@ def read_monomer_qmeta(self):
     def read_baits(self):
         con = sqlite3.connect(self.infile)
 
+        if not check_sqlite_table(con, 'NODE'):
+            sys.exit("Error: Your experimental design is not supported. At least two conditions are necessary for differential analysis. Switch 'level' to 'interaction' for visualization.")
+
         if self.combined:
             table = 'NODE'
         else:

secat/preprocess.py

@@ -65,16 +65,25 @@ def __init__(self, mitabfile):
 
     def read(self, mitabfile):
         def _extract_uniprotkb(string):
-            return [u for u in string.split("|") if "uniprotkb" in u][0].split("uniprotkb:")[1]
+            return [u for u in string.split("|") if "uniprotkb" in u][0].split("uniprotkb:")[1].split("-")[0]
 
         def _extract_score(string):
-            if 'score:' in string:
+            if 'intact-miscore:' in string:
+                return float([u for u in string.split("|") if "intact-miscore" in u][0].split("intact-miscore:")[1])
+            elif 'score:' in string:
                 return float([u for u in string.split("|") if "score" in u][0].split("score:")[1])
             else:
                 return 0
 
-        df = pd.read_csv(mitabfile, sep="\t", header = None, usecols=[0,1,14])
-        df.columns = ["bait_id","prey_id","interaction_confidence"]
+        df = pd.read_csv(mitabfile, sep="\t", header = None, usecols=[0,1,2,3,14])
+        df.columns = ["bait_id","prey_id","bait_id_alt","prey_id_alt","interaction_confidence"]
+
+        df['bait_id_alt'] = df['bait_id_alt'].replace(np.nan,'',regex=True)
+        df['prey_id_alt'] = df['prey_id_alt'].replace(np.nan,'',regex=True)
+
+        df['bait_id'] = df['bait_id'] + "|" + df['bait_id_alt']
+        df['prey_id'] = df['prey_id'] + "|" + df['prey_id_alt']
+        df = df.drop(columns=['bait_id_alt','prey_id_alt'])
 
         # Reduce DB to UniProtKB entries with scores
         df = df[df['bait_id'].str.contains('uniprotkb:') & df['prey_id'].str.contains('uniprotkb:') & (df['interaction_confidence'].str.contains('score:') | df['interaction_confidence'].str.contains('shortestPath:'))]
@@ -208,12 +217,12 @@ def identify(self, netfile):
 
     def unique_interactions(self, network):
         def get_interaction_id(x):
-            return '_'.join(sorted([x['bait_id'], x['prey_id']]))
+            return '__'.join(sorted([x['bait_id'], x['prey_id']]))
 
         network['interaction_id'] = network.apply(get_interaction_id, axis=1)
 
         network = network.groupby('interaction_id')['interaction_confidence'].max().reset_index()
-        network[['bait_id','prey_id']] = network.interaction_id.str.split('_', expand=True)
+        network[['bait_id','prey_id']] = network.interaction_id.str.split('__', expand=True)
         return network[['bait_id','prey_id','interaction_confidence']]
 
     def to_df(self):

secat/quantify.py

@@ -163,7 +163,7 @@ def contrast(self):
         con.close()
 
         if len(conditions) < 2:
-            sys.exit("Error: Your experimental design is not appropriate. At least two conditions are necessary for quantification.")
+            sys.exit("Error: Your experimental design is not supported. At least two conditions are necessary for differential analysis.")
 
         comparisons = []
         # prepare single-sample comparisons

setup.py

@@ -10,7 +10,7 @@
 
 setup(
     name='secat',
-    version='1.0.6',
+    version='1.0.7',
     description='Size-Exclusion Chromatography Algorithmic Toolkit',
     long_description=long_description,
     long_description_content_type='text/markdown',