def searchPfam(query, **kwargs):
"""Returns Pfam search results in a dictionary. Matching Pfam accession
as keys will map to evalue, alignment start and end residue positions.
:arg query: UniProt ID, PDB identifier, a protein sequence, or a sequence
file. Sequence queries must not contain without gaps and must be at
least 16 characters long
:type query: str
:arg timeout: timeout for blocking connection attempt in seconds, default
is 60
:type timeout: int
*query* can also be a PDB identifier, e.g. ``'1mkp'`` or ``'1mkpA'`` with
chain identifier. UniProt ID of the specified chain, or the first
protein chain will be used for searching the Pfam database."""
if isfile(query):
from prody.sequence import MSAFile
try:
seq = next(MSAFile(query))
except:
with openFile(query) as inp:
seq = ''.join(inp.read().split())
else:
seq = seq[0][1]
if not seq.isalpha():
raise ValueError('could not parse a sequence without gaps from ' +
query)
else:
seq = ''.join(query.split())
import xml.etree.cElementTree as ET
LOGGER.timeit('_pfam')
timeout = int(kwargs.get('timeout', 60))
if len(seq) >= MINSEQLEN:
if not seq.isalpha():
raise ValueError(repr(seq) + ' is not a valid sequence')
fseq = '>Seq\n' + seq
parameters = {'hmmdb': 'pfam', 'seq': fseq}
enc_params = urllib.urlencode(parameters).encode('utf-8')
request = urllib2.Request(
'https://www.ebi.ac.uk/Tools/hmmer/search/hmmscan', enc_params)
results_url = urllib2.urlopen(request).geturl()
#res_params = { 'output' : 'xml' }
res_params = {'format': 'tsv'}
enc_res_params = urllib.urlencode(res_params)
#modified_res_url = results_url + '?' + enc_res_params
modified_res_url = results_url.replace(
'results', 'download') + '?' + enc_res_params
result_request = urllib2.Request(modified_res_url)
# url = ( urllib2.urlopen(request).geturl() + '?output=xml')
LOGGER.debug('Submitted Pfam search for sequence "{0}...".'.format(
seq[:MINSEQLEN]))
try:
#xml = urllib2.urlopen(result_request).read()
tsv = urllib2.urlopen(result_request).read()
# openURL(url, timeout=timeout).read()
except:
raise ValueError('No matching Pfam domains were found.')
# try:
# root = ET.XML(xml)
# except Exception as err:
# raise ValueError('failed to parse results XML, check URL: ' + modified_res_url)
matches = {}
#for child in root[0]:
#if child.tag == 'hits':
# accession = child.get('acc')
# pfam_id = accession.split('.')[0]
# matches[pfam_id]={}
# matches[pfam_id]['accession']=accession
# matches[pfam_id]['class']='Domain'
# matches[pfam_id]['id']=child.get('name')
# matches[pfam_id]['locations']={}
# matches[pfam_id]['locations']['ali_end']=child[0].get('alisqto')
# matches[pfam_id]['locations']['ali_start']=child[0].get('alisqfrom')
# matches[pfam_id]['locations']['bitscore']=child[0].get('bitscore')
# matches[pfam_id]['locations']['end']=child[0].get('alisqto')
# matches[pfam_id]['locations']['evalue']=child.get('evalue')
# matches[pfam_id]['locations']['evidence']='hmmer v3.0'
# matches[pfam_id]['locations']['hmm_end']=child[0].get('alihmmto')
# matches[pfam_id]['locations']['hmm_start']=child[0].get('alihmmfrom')
# matches[pfam_id]['locations']['significant']=child[0].get('significant')
# matches[pfam_id]['locations']['start']=child[0].get('alisqfrom')
# matches[pfam_id]['type']='Pfam-A'
# return matches
if PY3K:
tsv = tsv.decode()
lines = tsv.split('\n')
keys = lines[0].split('\t')
root = {}
for i, line in enumerate(lines[1:-1]):
root[i] = {}
for j, key in enumerate(keys):
root[i][key] = line.split('\t')[j]
for child in root.values():
accession = child['Family Accession']
pfam_id = accession.split('.')[0]
matches[pfam_id] = {}
matches[pfam_id]['accession'] = accession
matches[pfam_id]['class'] = 'Domain'
matches[pfam_id]['id'] = child['Family id']
matches[pfam_id]['locations'] = {}
matches[pfam_id]['locations']['ali_end'] = child['Ali. End']
matches[pfam_id]['locations']['ali_start'] = child['Ali. Start']
matches[pfam_id]['locations']['bitscore'] = child['Bit Score']
matches[pfam_id]['locations']['end'] = child['Env. End']
matches[pfam_id]['locations']['cond_evalue'] = child[
'Cond. E-value']
matches[pfam_id]['locations']['ind_evalue'] = child['Ind. E-value']
matches[pfam_id]['locations']['evidence'] = 'hmmer v3.0'
matches[pfam_id]['locations']['hmm_end'] = child['Model End']
matches[pfam_id]['locations']['hmm_start'] = child['Model Start']
#matches[pfam_id]['locations']['significant'] = child['significant']
matches[pfam_id]['locations']['start'] = child['Env. Start']
matches[pfam_id]['type'] = 'Pfam-A'
return matches
else:
if len(seq) <= 5:
idcode = None
from prody import parsePDBHeader
try:
polymers = parsePDBHeader(seq[:4], 'polymers')
except Exception as err:
LOGGER.warn('failed to parse header for {0} ({1})'.format(
seq[:4], str(err)))
else:
chid = seq[4:].upper()
for poly in polymers:
if chid and poly.chid != chid:
continue
for dbref in poly.dbrefs:
if dbref.database != 'UniProt':
continue
idcode = dbref.idcode
accession = dbref.accession
LOGGER.info('UniProt ID code {0} for {1} chain '
'{2} will be used.'.format(
idcode, seq[:4], poly.chid))
break
if idcode is not None:
break
if idcode is None:
LOGGER.warn('A UniProt ID code for PDB {0} could not be '
'parsed.'.format(repr(seq)))
url = prefix + 'protein/' + seq + '?output=xml'
else:
url = prefix + 'protein/' + idcode + '?output=xml'
else:
url = prefix + 'protein/' + seq + '?output=xml'
LOGGER.debug('Retrieving Pfam search results: ' + url)
xml = None
while LOGGER.timing('_pfam') < timeout:
try:
xml = openURL(url, timeout=timeout).read()
except Exception:
pass
else:
if xml not in ['PEND', 'RUN']:
break
if not xml:
raise IOError('Pfam search timed out or failed to parse results '
'XML, check URL: ' + url)
else:
LOGGER.report('Pfam search completed in %.2fs.', '_pfam')
if xml.find(b'There was a system error on your last request.') > 0:
LOGGER.warn('No Pfam matches found for: ' + seq)
return None
elif xml.find(b'No valid UniProt accession or ID') > 0:
try:
url = prefix + 'protein/' + accession + '?output=xml'
xml = openURL(url, timeout=timeout).read()
except:
try:
ag = parsePDB(seq, subset='ca')
ag_seq = ag.getSequence()
return searchPfam(ag_seq)
except:
raise ValueError('No valid UniProt accession or ID for: ' +
seq)
try:
root = ET.XML(xml)
except Exception as err:
raise ValueError('failed to parse results XML, check URL: ' + url)
if len(seq) >= MINSEQLEN:
try:
xml_matches = root[0][0][0][0]
except IndexError:
raise ValueError('failed to parse results XML, check URL: ' + url)
else:
key = '{' + prefix + '}'
results = dictElement(root[0], key)
try:
xml_matches = results['matches']
except KeyError:
raise ValueError('failed to parse results XML, check URL: ' + url)
matches = dict()
for child in xml_matches:
try:
accession = child.attrib['accession'][:7]
except KeyError:
raise ValueError('failed to parse results XML, check URL: ' + url)
if not re.search('^P(F|B)[0-9]{5}$', accession):
raise ValueError('{0} does not match pfam accession'
' format'.format(accession))
match = matches.setdefault(accession, dict(child.items()))
locations = match.setdefault('locations', [])
for loc in child:
locations.append(dict(loc.items()))
if len(seq) < MINSEQLEN:
query = 'Query ' + repr(query)
else:
query = 'Query sequence'
if matches:
LOGGER.info(query + ' matched {0} Pfam families.'.format(len(matches)))
else:
LOGGER.info(query + ' did not match any Pfam families.')
return matches
def searchPfam(query, **kwargs):
"""Return Pfam search results in a dictionary. Matching Pfam accession
as keys will map to evalue, alignment start and end residue positions.
:arg query: UniProt ID, PDB identifier, protein sequence, or a sequence
file, sequence queries must not contain without gaps and must be at
least 16 characters long
:type query: str
:arg timeout: timeout for blocking connection attempt in seconds, default
is 60
:type timeout: int
*query* can also be a PDB identifier, e.g. ``'1mkp'`` or ``'1mkpA'`` with
chain identifier. UniProt ID of the specified chain, or the first
protein chain will be used for searching the Pfam database."""
prefix = '{http://pfam.xfam.org/}'
query = str(query)
if isfile(query):
from prody.sequence import MSAFile
try:
seq = next(MSAFile(query))
except:
with openFile(query) as inp:
seq = ''.join(inp.read().split())
else:
seq = seq[0][1]
if not seq.isalpha():
raise ValueError('could not parse a sequence without gaps from ' +
query)
else:
seq = ''.join(query.split())
import xml.etree.cElementTree as ET
LOGGER.timeit('_pfam')
timeout = int(kwargs.get('timeout', 60))
if len(seq) >= MINSEQLEN:
if not seq.isalpha():
raise ValueError(repr(seq) + ' is not a valid sequence')
fseq = '>Seq\n' + seq
parameters = { 'hmmdb' : 'pfam', 'seq': fseq }
enc_params = urllib.urlencode(parameters)
request = urllib2.Request('http://hmmer.janelia.org/search/hmmscan', enc_params)
url = ( urllib2.urlopen(request).geturl() + '?output=xml')
LOGGER.debug('Submitted Pfam search for sequence "{0}...".'
.format(seq[:MINSEQLEN]))
xml = openURL(url, timeout=timeout).read()
try:
root = ET.XML(xml)
except Exception as err:
raise ValueError('failed to parse results XML, check URL: ' + url)
matches = {}
for child in root[0]:
if child.tag == 'hits':
accession = child.get('acc')
pfam_id = accession.split('.')[0]
matches[pfam_id]={}
matches[pfam_id]['accession']=accession
matches[pfam_id]['class']='Domain'
matches[pfam_id]['id']=child.get('name')
matches[pfam_id]['locations']={}
matches[pfam_id]['locations']['ali_end']=child[0].get('alisqto')
matches[pfam_id]['locations']['ali_start']=child[0].get('alisqfrom')
matches[pfam_id]['locations']['bitscore']=child[0].get('bitscore')
matches[pfam_id]['locations']['end']=child[0].get('alisqto')
matches[pfam_id]['locations']['evalue']=child.get('evalue')
matches[pfam_id]['locations']['evidence']='hmmer v3.0'
matches[pfam_id]['locations']['hmm_end']=child[0].get('alihmmto')
matches[pfam_id]['locations']['hmm_start']=child[0].get('alihmmfrom')
matches[pfam_id]['locations']['significant']=child[0].get('significant')
matches[pfam_id]['locations']['start']=child[0].get('alisqfrom')
matches[pfam_id]['type']='Pfam-A'
return matches
else:
if len(seq) <= 5:
idcode = None
from prody import parsePDBHeader
try:
polymers = parsePDBHeader(seq[:4], 'polymers')
except Exception as err:
LOGGER.warn('failed to parse header for {0} ({1})'
.format(seq[:4], str(err)))
else:
chid = seq[4:].upper()
for poly in polymers:
if chid and poly.chid != chid:
continue
for dbref in poly.dbrefs:
if dbref.database != 'UniProt':
continue
idcode = dbref.idcode
LOGGER.info('UniProt ID code {0} for {1} chain '
'{2} will be used.'
.format(idcode, seq[:4], poly.chid))
break
if idcode is not None:
break
if idcode is None:
LOGGER.warn('A UniProt ID code for PDB {0} could not be '
'parsed.'.format(repr(seq)))
url = 'http://pfam.xfam.org/protein/' + seq + '?output=xml'
else:
url = ('http://pfam.xfam.org/protein/' +
idcode + '?output=xml')
else:
url = 'http://pfam.xfam.org/protein/' + seq + '?output=xml'
LOGGER.debug('Retrieving Pfam search results: ' + url)
xml = None
while LOGGER.timing('_pfam') < timeout:
try:
xml = openURL(url, timeout=timeout).read()
except Exception:
pass
else:
if xml:
break
if not xml:
raise IOError('Pfam search timed out or failed to parse results '
'XML, check URL: ' + url)
else:
LOGGER.report('Pfam search completed in %.2fs.', '_pfam')
if xml.find(b'There was a system error on your last request.') > 0:
LOGGER.warn('No Pfam matches found for: ' + seq)
return None
try:
root = ET.XML(xml)
except Exception as err:
raise ValueError('failed to parse results XML, check URL: ' + url)
if len(seq) >= MINSEQLEN:
try:
xml_matches = root[0][0][0][0]
except IndexError:
raise ValueError('failed to parse results XML, check URL: ' + url)
else:
results = dictElement(root[0], prefix)
try:
xml_matches = results['matches']
except KeyError:
raise ValueError('failed to parse results XML, check URL: ' + url)
matches = dict()
for child in xml_matches:
try:
accession = child.attrib['accession'][:7]
except KeyError:
raise ValueError('failed to parse results XML, check URL: ' + url)
if not re.search('^P(F|B)[0-9]{5}$', accession):
raise ValueError('{0} does not match pfam accession'
' format'.format(accession))
match = matches.setdefault(accession, dict(child.items()))
locations = match.setdefault('locations', [])
for loc in child:
locations.append(dict(loc.items()))
if len(seq) < MINSEQLEN:
query = 'Query ' + repr(query)
else:
query = 'Query sequence'
if matches:
LOGGER.info(query + ' matched {0} Pfam families.'.format(len(matches)))
else:
LOGGER.info(query + ' did not match any Pfam families.')
return matches
def searchPfam(query, search_b=False, skip_a=False, **kwargs):
"""Return Pfam search results in a dictionary. Matching Pfam accession
as keys will map to evalue, alignment start and end residue positions.
:arg query: UniProt ID, PDB identifier, protein sequence, or a sequence
file, sequence queries must not contain without gaps and must be at
least 16 characters long
:type query: str
:arg search_b: search Pfam-B families when **True**
:type search_b: bool
:arg skip_a: do not search Pfam-A families when **True**
:type skip_a: bool
:arg ga: use gathering threshold when **True**
:type ga: bool
:arg evalue: user specified e-value cutoff, must be smaller than 10.0
:type evalue: float
:arg timeout: timeout for blocking connection attempt in seconds, default
is 60
:type timeout: int
*query* can also be a PDB identifier, e.g. ``'1mkp'`` or ``'1mkpA'`` with
chain identifier. UniProt ID of the specified chain, or the first
protein chain will be used for searching the Pfam database."""
prefix = '{http://pfam.sanger.ac.uk/}'
query = str(query)
if isfile(query):
from prody.sequence import MSAFile
try:
seq = next(MSAFile(query))
except:
with openFile(query) as inp:
seq = ''.join(inp.read().split())
else:
seq = seq[0][1]
if not seq.isalpha():
raise ValueError('could not parse a sequence without gaps from ' +
query)
else:
seq = ''.join(query.split())
import xml.etree.cElementTree as ET
LOGGER.timeit('_pfam')
timeout = int(kwargs.get('timeout', 60))
if len(seq) >= MINSEQLEN:
if not seq.isalpha():
raise ValueError(repr(seq) + ' is not a valid sequence')
urlextension = ''
if kwargs:
ga = int(kwargs.get('ga', 1))
if not (ga == 1 or ga == 0):
raise ValueError('ga must be either 0 or 1')
evalue = kwargs.get('evalue', None)
if evalue:
if not float(evalue) <= 10.0:
raise ValueError('evalue must be a valid float < 10.0')
urlextension = urlextension + '&evalue=' + str(evalue)
else:
urlextension = urlextension + '&ga=' + str(ga)
search_b = int(bool(search_b))
skip_a = int(bool(skip_a))
if skip_a == 1:
search_b = 1
urlextension = urlextension + '&searchBs=' + str(search_b)
urlextension = urlextension + '&skipAs='******'http://pfam.sanger.ac.uk/search/sequence?seq=' + str(seq) +
urlextension + '&output=xml')
LOGGER.debug('Submitted Pfam search for sequence "{0}...".'
.format(seq[:MINSEQLEN]))
xml = openURL(url, timeout=timeout).read()
try:
root = ET.XML(xml)
except Exception as err:
raise ValueError('failed to parse results XML, check URL: ' + url)
try:
url = dictElement(root[0], prefix)['result_url']
except (IndexError, KeyError):
raise ValueError('failed to parse results XML, check URL: ' + url)
else:
if len(seq) <= 5:
idcode = None
from prody import parsePDBHeader
try:
polymers = parsePDBHeader(seq[:4], 'polymers')
except Exception as err:
LOGGER.warn('failed to parse header for {0} ({1})'
.format(seq[:4], str(err)))
else:
chid = seq[4:].upper()
for poly in polymers:
if chid and poly.chid != chid:
continue
for dbref in poly.dbrefs:
if dbref.database != 'UniProt':
continue
idcode = dbref.idcode
LOGGER.info('UniProt ID code {0} for {1} chain '
'{2} will be used.'
.format(idcode, seq[:4], poly.chid))
break
if idcode is not None:
break
if idcode is None:
LOGGER.warn('A UniProt ID code for PDB {0} could not be '
'parsed.'.format(repr(seq)))
url = 'http://pfam.sanger.ac.uk/protein/' + seq + '?output=xml'
else:
url = ('http://pfam.sanger.ac.uk/protein/' +
idcode + '?output=xml')
else:
url = 'http://pfam.sanger.ac.uk/protein/' + seq + '?output=xml'
LOGGER.debug('Retrieving Pfam search results: ' + url)
xml = None
while LOGGER.timing('_pfam') < timeout:
try:
xml = openURL(url, timeout=timeout).read()
except Exception:
pass
#else:
# if xml:
# break
if not xml:
raise IOError('Pfam search timed out or failed to parse results '
'XML, check URL: ' + url)
else:
LOGGER.report('Pfam search completed in %.2fs.', '_pfam')
if xml.find(b'There was a system error on your last request.') > 0:
LOGGER.warn('No Pfam matches found for: ' + seq)
return None
try:
root = ET.XML(xml)
except Exception as err:
raise ValueError('failed to parse results XML, check URL: ' + url)
if len(seq) >= MINSEQLEN:
try:
xml_matches = root[0][0][0][0]
except IndexError:
raise ValueError('failed to parse results XML, check URL: ' + url)
else:
results = dictElement(root[0], prefix)
try:
xml_matches = results['matches']
except KeyError:
raise ValueError('failed to parse results XML, check URL: ' + url)
matches = dict()
for child in xml_matches:
try:
accession = child.attrib['accession'][:7]
except KeyError:
raise ValueError('failed to parse results XML, check URL: ' + url)
if not re.search('^P(F|B)[0-9]{5}$', accession):
raise ValueError('{0} does not match pfam accession'
' format'.format(accession))
match = matches.setdefault(accession, dict(child.items()))
locations = match.setdefault('locations', [])
for loc in child:
locations.append(dict(loc.items()))
if len(seq) < MINSEQLEN:
query = 'Query ' + repr(query)
else:
query = 'Query sequence'
if matches:
LOGGER.info(query + ' matched {0} Pfam families.'.format(len(matches)))
else:
LOGGER.info(query + ' did not match any Pfam families.')
return matches