def translate_seq(sequence, strand):
'''
:param sequence: DNA sequence
:param strand: strand of DNA sequence
:return: translated DNA sequence
'''
if str(strand) == "1":
return standard_code.translate(sequence)
else:
return standard_code.translate(reverse_complement(sequence))
def ensembl_construct_sequences(psm_hash,ensembl,transcript_ids,database_v,species,three_frame_translation,mode,):
'''
:param psm_hash: dictionair with protein / ensembl information ( see prepareAnnotationENSEMBL)
:param ensembl:ensembl genome
:param transcript_ids: list of transcrip ids (converted from protein IDs)
:param database_v: database version
:param species: species name
:return: dictionairy mapping proteins into ENSEMBL
'''
print "Commencing transcript and protein sequence retrieval"
no_protein_seq=[]
biomart_key_hash={}
stable_transcript_ids=[]
for key in psm_hash.keys():
biomart_key_hash[psm_hash[key]['transcript_id']]=key
stable_transcript_ids.append(psm_hash[key]['transcript_id'])
# Retrieve cds,chr,transcript_id and strand from biomart
biomart_result=proBAM_biomart.retrieve_data_from_biomart(database_v,species,stable_transcript_ids,three_frame_translation)
for row in biomart_result:
row=row.split("\t")
try:
psm_hash[biomart_key_hash[row[1]]]['transcript_seq']=row[0]
psm_hash[biomart_key_hash[row[1]]]['protein_seq']=standard_code.translate(row[0])
#TODO what to do with "special" ensembl chromosomes: currently leave them out => bam conversion
#TODO considers these psms unmapped
#if "_" in row[2]:
# print row[1],row[2]
psm_hash[biomart_key_hash[row[1]]]['chr']=row[2]
psm_hash[biomart_key_hash[row[1]]]['strand']=row[3]
del row
except IndexError:
pass
del biomart_result
# get exons directly from core database
temp_exon_hash=get_ensembl_exons(ensembl,transcript_ids,psm_hash,mode)
exon_hash=temp_exon_hash[0]
psm_hash=temp_exon_hash[1]
del temp_exon_hash
# retrieve protein sequences for transcript where the protein sequence could not be fetched automatically
for key in no_protein_seq:
psm_hash[key]['transcript_seq']=retrieve_protein_seq(psm_hash[key]['transcript_seq'],
exon_hash[psm_hash[key]['transcript_id']],
psm_hash[key]['5UTR_offset'],
psm_hash[key]['start_exon_rank'])
#translate till stop codon
psm_hash[key]['protein_seq']=standard_code.translate(psm_hash[key]
['transcript_seq']).partition('*')[0]
return [psm_hash,exon_hash]
def map_peptide_to_protein_3frame(peptide_seq, transcript_seq,
allowed_mismatches, strand):
'''
:param peptide_seq: peptide sequence (string)
:param transcript_seq: transcript sequence (string)
:param allowed_mismatches: number of allowed mismatches
:param strand: chromosome strand
:return: number of hits of peptide on protein
'''
size_adjust = -1 # adjust size of transcript for starting at +1/+2frame
hits = []
pre_post_aa = ['', '']
pep_length = len(peptide_seq)
frame = [0] * 3
frame[0] = standard_code.translate(transcript_seq)
frame[1] = standard_code.translate(transcript_seq[1:])
frame[2] = standard_code.translate(transcript_seq[2:])
for f in frame:
size_adjust += 1
for i in range(0, (len(f) - pep_length)):
if hamming(peptide_seq, f[i:pep_length + i]) <= allowed_mismatches:
adjusted_hit_pos = (i * 3) + size_adjust
hits.append([
adjusted_hit_pos,
hamming(peptide_seq, f[i:pep_length + i])
])
# compute 2 preceding AA
if (i - 1) == 0:
pre_post_aa[0] = f[(i - 1)]
elif (i - 2) >= 0:
pre_post_aa[0] = f[(i - 2):i]
else:
pre_post_aa[0] = "*"
# compute 2 folowwing AA
if (i + 1) == (len(f) - 1):
pre_post_aa[1] = f[pep_length + i]
elif (i + 2) <= (len(f) - 1):
pre_post_aa[1] = f[(pep_length + i):(pep_length + i + 2)]
else:
pre_post_aa[1] = "*"
return [hits, pre_post_aa]
def map_peptide_to_protein_3frame(peptide_seq,transcript_seq,allowed_mismatches,strand):
'''
:param peptide_seq: peptide sequence (string)
:param transcript_seq: transcript sequence (string)
:param allowed_mismatches: number of allowed mismatches
:param strand: chromosome strand
:return: number of hits of peptide on protein
'''
size_adjust=-1 # adjust size of transcript for starting at +1/+2frame
hits=[]
pre_post_aa=['','']
pep_length=len(peptide_seq)
frame=[0]*3
frame[0]=standard_code.translate(transcript_seq)
frame[1]=standard_code.translate(transcript_seq[1:])
frame[2]=standard_code.translate(transcript_seq[2:])
for f in frame:
size_adjust+=1
for i in range(0,(len(f)-pep_length)):
if hamming(peptide_seq,f[i:pep_length+i]) <= allowed_mismatches:
adjusted_hit_pos=(i*3)+size_adjust
hits.append([adjusted_hit_pos,hamming(peptide_seq,f[i:pep_length+i])])
# compute 2 preceding AA
if (i - 1) == 0:
pre_post_aa[0] = f[(i - 1)]
elif (i - 2) >= 0:
pre_post_aa[0] = f[(i - 2):i]
else:
pre_post_aa[0] = "*"
# compute 2 folowwing AA
if (i + 1) == (len(f) - 1):
pre_post_aa[1] = f[pep_length + i]
elif (i + 2) <= (len(f) - 1):
pre_post_aa[1] = f[(pep_length + i):(pep_length + i + 2)]
else:
pre_post_aa[1] = "*"
return [hits,pre_post_aa]
def ensembl_construct_sequences(
psm_hash,
mysql_db,
transcript_ids,
database_v,
species,
three_frame_translation,
mode,
):
'''
:param psm_hash: dictionair with protein / ensembl information ( see prepareAnnotationENSEMBL)
:param ensembl:ensembl genome
:param transcript_ids: list of transcrip ids (converted from protein IDs)
:param database_v: database version
:param species: species name
:return: dictionairy mapping proteins into ENSEMBL
'''
print "Commencing transcript and protein sequence retrieval"
no_protein_seq = []
biomart_key_hash = {}
stable_transcript_ids = []
for key in psm_hash.keys():
biomart_key_hash[psm_hash[key]['transcript_id']] = key
stable_transcript_ids.append(psm_hash[key]['transcript_id'])
chunked_stable_transcript_id = chunkIt(stable_transcript_ids, 10)
process = 0
c = 0
for chunk in chunked_stable_transcript_id:
# Retrieve cds,chr,transcript_id and strand from biomart
try:
biomart_result = proBAM_biomart.retrieve_data_from_biomart(
database_v, species, chunk, three_frame_translation)
except AttributeError:
time.sleep(60)
print "BioMart connection timeout, reconnecting to BioMart"
biomart_result = proBAM_biomart.retrieve_data_from_biomart(
database_v, species, chunk, three_frame_translation)
for row in biomart_result:
row = row.split("\t")
try:
psm_hash[biomart_key_hash[row[1]]]['transcript_seq'] = row[0]
psm_hash[biomart_key_hash[row[1]]]['shift'] = _calc_seq_shift_(
row[0])
psm_hash[biomart_key_hash[
row[1]]]['protein_seq'] = standard_code.translate(row[0])
psm_hash[biomart_key_hash[row[1]]]['chr'] = row[2]
psm_hash[biomart_key_hash[row[1]]]['strand'] = row[3]
del row
except IndexError:
pass
del biomart_result
if process < 100:
process += 10
print str(process) + "% ",
print " "
# get exons directly from core database
temp_exon_hash = get_ensembl_exons(mysql_db, transcript_ids, psm_hash,
mode)
exon_hash = temp_exon_hash[0]
psm_hash = temp_exon_hash[1]
del temp_exon_hash
# retrieve protein sequences for transcript where the protein sequence could not be fetched automatically
for key in no_protein_seq:
psm_hash[key]['transcript_seq'] = retrieve_protein_seq(
psm_hash[key]['transcript_seq'],
exon_hash[psm_hash[key]['transcript_id']],
psm_hash[key]['5UTR_offset'], psm_hash[key]['start_exon_rank'])
psm_hash[key]['shift'] = _calc_seq_shift_(
psm_hash[key]['transcript_seq'])
#translate till stop codon
psm_hash[key]['protein_seq'] = standard_code.translate(
psm_hash[key]['transcript_seq']).partition('*')[0]
return [psm_hash, exon_hash]