def filter_whole_proteins(elements, protein_fasta, lookup, seqtype, ns,
deamidation, minpeplen, enforce_tryp):
whole_proteins = {
str(prot.seq).replace('L', 'I'): prot.id
for prot in fasta.parse_fasta(protein_fasta)
}
whole_proteins = {v: k for k, v in whole_proteins.items()}
for element in elements:
seq_matches_protein = False
element_seqs = get_seqs_from_element(element, seqtype, ns, deamidation)
element_prots = {
seq:
[(protid, pos)
for protid, pos in lookup.get_protein_from_pep(seq[:minpeplen])]
for seq in element_seqs
}
for pepseq, proteins in element_prots.items():
for prot_id, pos in proteins:
protseq = whole_proteins[prot_id]
if pepseq in protseq:
if enforce_tryp and (
pos == 0 or not set([pepseq[-1], protseq[pos - 1]
]).difference(['K', 'R'])):
# pepseq is tryptic on both ends, or
# pepseq is an N-term peptide),
# matches to protein seq so remove
seq_matches_protein = True
break
elif not enforce_tryp:
seq_matches_protein = True
break
if seq_matches_protein:
formatting.clear_el(element)
else:
yield formatting.string_and_clear(element, ns)
def filter_whole_proteins(elements, protein_fasta, lookup, seqtype, ns,
deamidation, minpeplen, enforce_tryp):
whole_proteins = {str(prot.seq).replace('L', 'I'): prot.id for prot in
fasta.parse_fasta(protein_fasta)}
whole_proteins = {v: k for k, v in whole_proteins.items()}
for element in elements:
seq_matches_protein = False
element_seqs = get_seqs_from_element(element, seqtype, ns, deamidation)
element_prots = {seq: [(protid, pos) for protid, pos in
lookup.get_protein_from_pep(seq[:minpeplen])]
for seq in element_seqs}
for pepseq, proteins in element_prots.items():
for prot_id, pos in proteins:
protseq = whole_proteins[prot_id]
if pepseq in protseq:
if enforce_tryp and (pos == 0 or not set(
[pepseq[-1],
protseq[pos - 1]]).difference(['K', 'R'])):
# pepseq is tryptic on both ends, or
# pepseq is an N-term peptide),
# matches to protein seq so remove
seq_matches_protein = True
break
elif not enforce_tryp:
seq_matches_protein = True
break
if seq_matches_protein:
formatting.clear_el(element)
else:
yield formatting.string_and_clear(element, ns)
def reassign_elements(elements, stats, ns):
for el in elements:
score = round(float(el.xpath('xmlns:svm_score',
namespaces=ns)[0].text), 5)
oldq = el.xpath('xmlns:q_value', namespaces=ns)[0]
oldpep = el.xpath('xmlns:pep', namespaces=ns)[0]
newq, newpep, warning = lookup_statistic(score, stats)
if warning is not None:
sys.stdout.write(warning)
oldq.text, oldpep.text = newq, newpep
yield formatting.string_and_clear(el, ns)
def filter_peptide_length(features, elementtype, ns, minlen=0, maxlen=None):
minlen = int(minlen)
if maxlen is None:
maxlen = float('inf')
else:
maxlen = int(maxlen)
for feat in features:
seq = get_either_seq(elementtype, feat, ns)
seq = strip_modifications(seq)
if len(seq) >= minlen and len(seq) <= maxlen:
yield formatting.string_and_clear(feat, ns)
else:
formatting.clear_el(feat)
def filter_peptide_length(features, elementtype, ns, minlen=0, maxlen=None):
minlen = int(minlen)
if maxlen is None:
maxlen = float('inf')
else:
maxlen = int(maxlen)
for feat in features:
seq = get_either_seq(elementtype, feat, ns)
seq = strip_modifications(seq)
if len(seq) >= minlen and len(seq) <= maxlen:
yield formatting.string_and_clear(feat, ns)
else:
formatting.clear_el(feat)
def protein_header_split_generator(elements, headers, ns):
"""Loop through proteins of each PSM/peptide. If a protein does not
match any of headers, discard PSM/peptide immediately"""
for el in elements:
header_not_matching = False
for protein in el.findall('{%s}protein_id' % ns['xmlns']):
if not any((re.search(h, protein.text) for h in headers)):
header_not_matching = True
break
if header_not_matching:
formatting.clear_el(el)
else:
yield formatting.string_and_clear(el, ns)
def protein_header_split_generator(elements, headers, ns):
"""Loop through proteins of each PSM/peptide. If a protein does not
match any of headers, discard PSM/peptide immediately"""
for el in elements:
header_not_matching = False
for protein in el.findall('{%s}protein_id' % ns['xmlns']):
if not any((re.search(h, protein.text) for h in headers)):
header_not_matching = True
break
if header_not_matching:
formatting.clear_el(el)
else:
yield formatting.string_and_clear(el, ns)
def merge_peptides(fns, ns):
"""Loops peptides from multiple files, fetches PSMs from
sequence:PSM map, outputs correctly PSM mapped peptides"""
peptides_to_map = reader.generate_peptides_multiple_fractions(fns, ns)
psmmap = create_merge_psm_map(peptides_to_map, ns)
peptides = reader.generate_peptides_multiple_fractions(fns, ns)
for peptide in peptides:
seq = reader.get_peptide_seq(peptide, ns)
psm_ids = reader.get_psm_ids_from_peptide(peptide, ns)
# remove current psm ids, repopulate with stored ones
psm_ids.clear()
for new_psm_id in psmmap[seq]:
etree.SubElement(psm_ids, 'psm_id').text = new_psm_id
yield formatting.string_and_clear(peptide, ns)
def merge_peptides(fns, ns):
"""Loops peptides from multiple files, fetches PSMs from
sequence:PSM map, outputs correctly PSM mapped peptides"""
peptides_to_map = reader.generate_peptides_multiple_fractions(fns, ns)
psmmap = create_merge_psm_map(peptides_to_map, ns)
peptides = reader.generate_peptides_multiple_fractions(fns, ns)
for peptide in peptides:
seq = reader.get_peptide_seq(peptide, ns)
psm_ids = reader.get_psm_ids_from_peptide(peptide, ns)
# remove current psm ids, repopulate with stored ones
psm_ids.clear()
for new_psm_id in psmmap[seq]:
etree.SubElement(psm_ids, 'psm_id').text = new_psm_id
yield formatting.string_and_clear(peptide, ns)
def filter_known_searchspace(elements, seqtype, lookup, ns, ntermwildcards,
deamidation):
"""Yields peptides from generator as long as their sequence is not found in
known search space dict. Useful for excluding peptides that are found in
e.g. ENSEMBL or similar"""
for element in elements:
seq_is_known = False
for seq in get_seqs_from_element(element, seqtype, ns, deamidation):
if lookup.check_seq_exists(seq, ntermwildcards):
seq_is_known = True
break
if seq_is_known:
formatting.clear_el(element)
else:
yield formatting.string_and_clear(element, ns)
def filter_known_searchspace(elements, seqtype, lookup, ns, ntermwildcards,
deamidation):
"""Yields peptides from generator as long as their sequence is not found in
known search space dict. Useful for excluding peptides that are found in
e.g. ENSEMBL or similar"""
for element in elements:
seq_is_known = False
for seq in get_seqs_from_element(element, seqtype, ns, deamidation):
if lookup.check_seq_exists(seq, ntermwildcards):
seq_is_known = True
break
if seq_is_known:
formatting.clear_el(element)
else:
yield formatting.string_and_clear(element, ns)
def protein_header_split_generator(elements, ns, can_headers, headers):
"""Loop through proteins of each PSM/peptide. If a protein does not
match any of headers, discard PSM/peptide immediately"""
for el in elements:
header_matching = False
can = False
for protein in el.findall('{%s}protein_id' % ns['xmlns']):
if any(re.search(h, protein.text) for h in can_headers):
can = True
break #as soon as a canonical match was found break
"""for classes other than known,
check if there is at least one protein matching the specified header
and those with matches to the canonical proteins will not be used"""
if any(re.search(h, protein.text) for h in headers):
header_matching = True
if (header_matching and not can) or ((headers == can_headers) and can):
yield formatting.string_and_clear(el, ns)
else:
formatting.clear_el(el)
def generate_tags_multiple_files_strings(input_files, ns, tag, ignore_tags):
"""
Creates stringified xml output of elements with certain tag.
"""
for el in generate_tags_multiple_files(input_files, tag, ignore_tags, ns):
yield formatting.string_and_clear(el, ns)
def target_decoy_generator(element_generator, decoy, ns):
for el in element_generator:
if el.attrib['{%s}decoy' % ns['xmlns']] == decoy:
yield formatting.string_and_clear(el, ns)
else:
formatting.clear_el(el)
def generate_tags_multiple_files_strings(input_files, ns, tag, ignore_tags):
"""
Creates stringified xml output of elements with certain tag.
"""
for el in generate_tags_multiple_files(input_files, tag, ignore_tags, ns):
yield formatting.string_and_clear(el, ns)
def target_decoy_generator(element_generator, decoy, ns):
for el in element_generator:
if el.attrib['{%s}decoy' % ns['xmlns']] == decoy:
yield formatting.string_and_clear(el, ns)
else:
formatting.clear_el(el)