Skip to content

Commit

Permalink
enhancing BRENDA parsing
Browse files Browse the repository at this point in the history
  • Loading branch information
@jonrkarr
jonrkarr committed Apr 23, 2020
1 parent 648dbbf commit 48b933b
Showing 1 changed file with 130 additions and 36 deletions.
166 changes: 130 additions & 36 deletions datanator/data_source/brenda/core.py
View file
Original file line number Diff line number Diff line change
Expand Up @@ -15,13 +15,14 @@

import ete3
import pickle
import os.path
import re
import warnings


class Brenda(object):
RAW_FILENAME = './brenda_download.txt'
PROCESSED_FILENAME = './brenda.pkl'
RAW_FILENAME = os.path.join(os.path.dirname(__file__), 'brenda_download.txt')
PROCESSED_FILENAME = os.path.join(os.path.dirname(__file__), 'brenda.pkl')

LINE_CODES = {
'AC': 'ACTIVATING_COMPOUND',
Expand Down Expand Up @@ -150,6 +151,12 @@ def run(self, raw_filename=None, processed_filename=None):
for loc in enz['subcellular_localizations']:
loc['refs'] = [ec_data['refs'][ref_id] for ref_id in loc['ref_ids']]

for rxn in ec_data['natural_reactions']:
rxn['refs'] = [ec_data['refs'][ref_id] for ref_id in rxn['ref_ids']]

for rxn in ec_data['reactions']:
rxn['refs'] = [ec_data['refs'][ref_id] for ref_id in rxn['ref_ids']]

for k_cat in ec_data['k_cats']:
k_cat['refs'] = [ec_data['refs'][ref_id] for ref_id in k_cat['ref_ids']]

Expand All @@ -168,6 +175,12 @@ def run(self, raw_filename=None, processed_filename=None):
for loc in enz['subcellular_localizations']:
loc.pop('ref_ids')

for rxn in ec_data['natural_reactions']:
rxn.pop('ref_ids')

for rxn in ec_data['reactions']:
rxn.pop('ref_ids')

for k_cat in ec_data['k_cats']:
k_cat.pop('ref_ids')

Expand Down Expand Up @@ -196,6 +209,7 @@ def parse_ec_code(self, data, val):
ec_data = data[ec_code] = {
'name': None,
'systematic_name': None,
'natural_reactions': [],
'reactions': [],
'enzymes': {},
'k_cats': [],
Expand All @@ -216,7 +230,7 @@ def parse_content(self, ec_code, ec_data, type, val):
r'[ \n](.*?)'
r'([ \n]([A-Z][A-Z0-9\.]+)'
r'[ \n](GenBank|UniProt|SwissProt|))?'
r'([ \n]\(.*?\))?'
r'([ \n]\((.*?)\))?'
r'[ \n]<([0-9,\n]+)>$'
), val, re.DOTALL)

Expand All @@ -225,7 +239,8 @@ def parse_content(self, ec_code, ec_data, type, val):
taxon_id = self._ncbi_taxa.get_name_translator([taxon_name]).get(taxon_name, [None])[0]
xid = match.group(4) or None
namespace = match.group(5) or None
ref_ids = match.group(7).replace('\n', ',').strip().split(',')
comments = self.parse_comments(match.group(7))
ref_ids = match.group(8).replace('\n', ',').strip().split(',')

if id not in ec_data['enzymes']:
ec_data['enzymes'][id] = {
Expand All @@ -234,6 +249,7 @@ def parse_content(self, ec_code, ec_data, type, val):
'taxon': {'name': taxon_name, 'id': taxon_id} if taxon_name else None,
'tissues': [],
'subcellular_localizations': [],
'comments': comments,
'ref_ids': ref_ids,
'refs': None,
}
Expand All @@ -248,6 +264,7 @@ def parse_content(self, ec_code, ec_data, type, val):
if not ec_data['enzymes'][id]['taxon'] and taxon_name:
ec_data['enzymes'][id]['taxon'] = {'name': taxon_name, 'id': taxon_id}

ec_data['enzymes'][id]['comments'] += comments
ec_data['enzymes'][id]['ref_ids'] = sorted(set(ec_data['enzymes'][id]['ref_ids'] + ref_ids))

elif type == 'RN':
Expand All @@ -257,88 +274,98 @@ def parse_content(self, ec_code, ec_data, type, val):
ec_data['systematic_name'] = val.replace('\n', ' ').strip()

elif type == 'ST':
match = re.match(r'^#(.*?)#[ \n](.*?)([ \n]\(.*?\))?[ \n]<([0-9,\n]+)>$', val, re.DOTALL)
match = re.match(r'^#(.*?)#[ \n](.*?)([ \n]\((.*?)\))?[ \n]<([0-9,\n]+)>$', val, re.DOTALL)

enz_ids = match.group(1).replace('\n', ',').strip().split(',')
tissue = match.group(2).strip()
ref_ids = match.group(4).replace('\n', ',').strip().split(',')
comments = self.parse_comments(match.group(4))
ref_ids = match.group(5).replace('\n', ',').strip().split(',')
for enz_id in enz_ids:
enzyme = ec_data['enzymes'].get(enz_id, None)
if enzyme:
enzyme['tissues'].append({
'name': tissue,
'comments': self.filter_comments(comments, enz_id),
'ref_ids': ref_ids,
'refs': None,
})
else:
warnings.warn('{} does not have enzyme with id {}. Error due to {}'.format(ec_code, enz_id, val), UserWarning)

elif type == 'LO':
match = re.match(r'^#(.*?)#[ \n](.*?)([ \n]\(.*?\))?[ \n]<([0-9,\n]+)>$', val, re.DOTALL)
match = re.match(r'^#(.*?)#[ \n](.*?)([ \n]\((.*?)\))?[ \n]<([0-9,\n]+)>$', val, re.DOTALL)

enz_ids = match.group(1).replace('\n', ',').strip().split(',')
localization = match.group(2).strip()
ref_ids = match.group(4).replace('\n', ',').strip().split(',')
comments = self.parse_comments(match.group(4))
ref_ids = match.group(5).replace('\n', ',').strip().split(',')
for enz_id in enz_ids:
enzyme = ec_data['enzymes'].get(enz_id, None)
if enzyme:
ec_data['enzymes'][enz_id]['subcellular_localizations'].append({
'name': localization,
'comments': self.filter_comments(comments, enz_id),
'ref_ids': ref_ids,
'refs': None,
})
else:
warnings.warn('{} does not have enzyme with id {}. Error due to {}'.format(ec_code, enz_id, val), UserWarning)

elif type == 'SP':
match = re.match(r'^#(.*?)#[ \n](.*?)([ \n]\(.*?\))?([ \n]\|.*?\|)?([ \n]\{(r)\})?[ \n]<([0-9,\n]+)>$', val, re.DOTALL)
elif type == 'NSP':
match = re.match(r'^#(.*?)#[ \n](.*?)([ \n]\((.*?)\))?([ \n]\|.*?\|)?([ \n]\{(r|)\})?[ \n]<([0-9,\n]+)>$', val, re.DOTALL)
comments = self.parse_comments(match.group(4))
ref_ids = match.group(8).replace('\n', ',').strip().split(',')
ec_data['natural_reactions'].append({
'equation': match.group(2).replace('\n', ' ').strip(),
'reversible': match.group(7) == 'r',
'enz_ids': match.group(1).replace('\n', ',').strip().split(','),
'comments': comments,
'ref_ids': ref_ids,
'refs': None,
})

elif type == 'SP':
match = re.match(r'^#(.*?)#[ \n](.*?)([ \n]\((.*?)\))?([ \n]\|.*?\|)?([ \n]\{(r|)\})?[ \n]<([0-9,\n]+)>$', val, re.DOTALL)
comments = self.parse_comments(match.group(4))
ref_ids = match.group(8).replace('\n', ',').strip().split(',')
ec_data['reactions'].append({
'equation': match.group(2).replace('\n', ' ').strip(),
'reversible': match.group(6) == 'r',
'enzymes': match.group(1).replace('\n', ',').strip().split(','),
'reversible': match.group(7) == 'r',
'enz_ids': match.group(1).replace('\n', ',').strip().split(','),
'comments': comments,
'ref_ids': ref_ids,
'refs': None,
})

elif type in ['TN', 'KM']:
match = re.match(
'^#(.*?)#[ \n]((\d+(\.\d+)?)?-?(\d+(\.\d+)?)?)[ \n]{(.*?)}[ \n]([ \n]\((.*?)\))?[ \n]?<([0-9,\n]+)>$', val, re.DOTALL)

enz_ids = match.group(1).replace('\n', ',').split(',')
ref_ids = match.group(10).replace('\n', ',').strip().split(',')
comments = self.parse_comments(match.group(9))

datum = {
'substrate': match.group(7),
'value': match.group(2).replace('\n', '-'),
'value': match.group(2).replace('\n', ''),
'units': 's^-1' if type == 'TN' else 'mM',
'enzymes': [],
'wild_type': None,
'genetic_variant': None,
'temperature': None,
'ph': None,
'ref_ids': match.group(10).replace('\n', ',').strip().split(','),
'ref_ids': ref_ids,
'refs': None,
}

enz_ids = match.group(1).replace('\n', ',').split(',')
for enz_id in enz_ids:
enzyme = ec_data['enzymes'].get(enz_id, None)
if enzyme:
datum['enzymes'].append(enzyme)
else:
warnings.warn('{} does not have enzyme with id {}. Error due to {}'.format(ec_code, enz_id, val), UserWarning)

comments = match.group(9)
if comments:
if 'wild-type' in comments or 'native' in comments:
datum['wild_type'] = True

if 'mutant' in comments:
datum['genetic_variant'] = True
filtered_comments = self.filter_comments(comments, enz_id)

match = re.search('(\d+(\.\d+)?)°C', comments)
if match:
datum['temperature'] = float(match.group(1))
datum['enzymes'].append({
'enzyme': enzyme,
'comments': filtered_comments,
})

match = re.search('pH[ \n](\d+(\.\d+)?)', comments, re.DOTALL)
if match:
datum['ph'] = float(match.group(1))
else:
warnings.warn('{} does not have enzyme with id {}. Error due to {}'.format(ec_code, enz_id, val), UserWarning)

if type == 'TN':
ec_data['k_cats'].append(datum)
Expand Down Expand Up @@ -367,6 +394,73 @@ def parse_content(self, ec_code, ec_data, type, val):
'comments': match.group(6).replace('\n', ' ') if match.group(6) else None,
}

def parse_comments(self, comment_strs):
if not comment_strs:
return []

comment_strs = re.split(r';[ \n]', comment_strs.strip())
comments = []
for comment_str in comment_strs:
comment_match = re.match(r'^#([0-9,\n]+)#[ \n](.*?)[ \n]<([0-9,\n]+)>$', comment_str.strip())
if comment_match:
enz_ids = comment_match.group(1).replace('\n', ',').strip().split(',')
ref_ids = comment_match.group(3).replace('\n', ',').strip().split(',')

text = comment_match.group(2).replace('\n', ' ')
if 'wild-type' in text or 'native' in text:
wild_type = True
else:
wild_type = None

if 'mutant' in text:
genetic_variant = True
else:
genetic_variant = None

match = re.search(r'(^|,[ \n])(\d+(\.\d+)?)°C(,[ \n]|$)', text)
if match:
temperature = float(match.group(2))
else:
temperature = None

match = re.search(r'(^|,[ \n])pH[ \n](\d+(\.\d+)?)(,[ \n]|$)', text, re.DOTALL)
if match:
ph = float(match.group(2))
else:
ph = None

comments.append({
'enz_ref_ids': list(zip(enz_ids, ref_ids)),
'text': text,
'wild_type': wild_type,
'genetic_variant': genetic_variant,
'temperature': temperature,
'ph': ph,
})
else:
comments.append({
'enz_ref_ids': [(None, None)],
'text': comment_str,
'wild_type': None,
'genetic_variant': None,
'temperature': None,
'ph': None,
})

return comments

def filter_comments(self, comments, enz_id):
filtered_comments = []
for comment in comments:
has_comment = False
for enz_ref_id in comment['enz_ref_ids']:
if enz_ref_id[0] == enz_id:
has_comment = True
break
if has_comment:
filtered_comments.append(comment)
return filtered_comments


"""
names = set()
Expand Down

0 comments on commit 48b933b

Please sign in to comment.