From b7f17ea1bcc704dc3bdcaecd7286b5fedb300151 Mon Sep 17 00:00:00 2001
From: Alexey Strokach <alex.strokach@utoronto.ca>
Date: Wed, 15 Apr 2015 20:32:50 -0400
Subject: [PATCH] Creating the provean supporting set can fail if the protein
 starts with a weird amino acid (e.g. 'X', etc.). Fixed.

---
 elaspic/domain_alignment.py | 17 ++++++++++++++---
 elaspic/helper_functions.py |  1 +
 tests/test_database.py      |  2 +-
 3 files changed, 16 insertions(+), 4 deletions(-)

diff --git a/elaspic/domain_alignment.py b/elaspic/domain_alignment.py
index 67ffdda..0774ca4 100755
--- a/elaspic/domain_alignment.py
+++ b/elaspic/domain_alignment.py
@@ -160,8 +160,11 @@ def build_provean_supporting_set(
     """
     """
     # Get the required parameters
-    first_aa = uniprot_sequence[0]
-    domain_mutation = '{0}1{0}'.format(first_aa)
+    any_position = 0
+    while uniprot_sequence[any_position] not in hf.canonical_amino_acids:
+        any_position += 1
+    first_aa = uniprot_sequence[any_position]
+    domain_mutation = '{0}{1}{0}'.format(first_aa, any_position+1)
 
     uniprot_seqrecord = SeqRecord(
         seq=Seq(uniprot_sequence), id=str(uniprot_id), description=uniprot_name)
@@ -182,10 +185,18 @@ def build_provean_supporting_set(
         logger.error(error_message)
         raise errors.ProveanError(error_message)
 
+    provean_supset_length = None
     for line in result.split('\n'):
         if 'Number of supporting sequences used:' in line:
             provean_supset_length = int(line.split()[-1])
-
+    if provean_supset_length is None:
+        logger.error('Provean supporting set length could not be estimated. This is an error!')
+        logger.error('Provean result: {}'.format(result))
+        logger.error('Provean error_message: {}'.format(error_message))
+        logger.error('Provean return_code: {}'.format(return_code))
+        logger.error('Uniprot sequence: {}'.format(uniprot_sequence))
+        logger.error('First amino acid: {}'.format(first_aa))
+        logger.error('Domain mutation: {}'.format(domain_mutation))
     return provean_supset_filename, provean_supset_length
 
 
diff --git a/elaspic/helper_functions.py b/elaspic/helper_functions.py
index 960a0cb..862b017 100755
--- a/elaspic/helper_functions.py
+++ b/elaspic/helper_functions.py
@@ -24,6 +24,7 @@
 
 
 #%%
+canonical_amino_acids = 'ARNDCEQGHILKMFPSTWYV'
 uppercase = 'ABCDEFGHIJKLMNOPQRSTUVWXYZ'
 logger = None
  
diff --git a/tests/test_database.py b/tests/test_database.py
index 28986db..52b55a9 100644
--- a/tests/test_database.py
+++ b/tests/test_database.py
@@ -55,7 +55,7 @@
 psql_command_template = r""" \
 copy {db_schema}.{table_name} \
 from '{organism_folder}/{table_name}.tsv' \
-with csv delimiter E'\t' null '\N' escape '\\'
+with csv delimiter E'\t' null '\N' escape '\\\\'
 """
 
 def _format_configs(configs, table_name):