def run_blastp(query, subject, outfile, evalue=1e-4):
"""
This function runs blastp
:param query: file that contains query sequences
:param subject: file that contains the sequences used for blasting
:param outfile: the path to the file that will contain blast results if the blastp
is successful
:param evalue: threshold value for blast results, default is 1e-4 or user defined
:return: returns the amount of blast hits which says if the blastp was successful
"""
cmd = blastp(query=query,
subject=subject,
evalue=evalue,
outfmt=6,
out=outfile)
#print(cmd)
stdout, stderr = cmd()
#if stderr:
# logging.warn("STDERR = ", stderr)
if not os.path.isfile(outfile):
raise Exception('No BLAST output')
if os.path.getsize(outfile) == 0:
return 0
return len(open(outfile).read().splitlines())
def blast_gene(seq, database):
tempfasta = open('temp.fasta', 'w')
SeqIO.write(seq, tempfasta, 'fasta')
tempfasta.close()
run = blastp(query='temp.fasta',
db=database,
num_descriptions=5,
num_threads=6,
outfmt=5,
out='temp.xml')
run()
result_handle = open('temp.xml')
result = NCBIXML.read(result_handle)
rets = []
for i in result.descriptions:
ttl = i.title
e = i.e
if 'Tfl|' in ttl:
species = 'T. flavus'
d = ttl[ttl.find('Tfl'):]
elif 'Pfu|' in ttl:
species = 'P. funiculosum'
d = ttl[ttl.find('Pfu'):]
elif 'PMAA_' in ttl:
species = 'T. marneffei'
d = ttl[ttl.find('PMAA'):]
else:
species = ttl[ttl.find('[') + 1:ttl.find(']')]
d = ttl[ttl.find('| ') + 1:ttl.find('[') - 1]
rets.append(species)
rets.append(d)
rets.append(str(e))
return rets
def Blast(type,protein_sequence,start, end, genomic_sequence):
result=[]
ORF=[]
M = re.search('M',protein_sequence)
if M:
query = protein_sequence[M.start():]
query=query+"*"
temp = open("temp.ORF", "w")
print >>temp, '>blasting'
print >>temp, query
temp.close()
cline=blastp(query="'temp.ORF'", db = "./Blast/DB.blast.txt", evalue=0.01, outfmt=5, out=type +".BLAST")
os.system(str(cline))
blast_out=open(type+".BLAST")
string=str(blast_out.read())
DEF=re.search("<Hit_def>((.*))</Hit_def>",string)
if DEF:
if DEF.group(1)=='L1':
real_start=start+M.start()+M.start()+M.start()
result.append(type)
result.append('L1')
L1_pre=genomic_sequence[(start+3*M.start()):int(end)]
splice='(C|T)(C|T)(A|C|G|T)(C|T)AG(A)TG'
spliced=re.search(splice,str(L1_pre))
if spliced:
start_L1 = int(spliced.start())+6
if start_L1 % 3 == 0:
if start_L1 > 600:
L1_post=L1_pre
result.append(str(start_L1)+".."+str(end))
result.append(str(L1_post))
result.append(Seq(str(L1_post)).translate())
else:
L1_post=L1_pre[start_L1:]
result.append(str(int(real_start+1)+int(start_L1))+".."+str(end))
result.append(str(L1_post))
result.append(Seq(str(L1_post)).translate())
else:
L1_post=L1_pre
result.append(str(real_start+1)+".."+str(end))
result.append(str(L1_post))
result.append(Seq(str(L1_post)).translate())
else:
L1_post=L1_pre
result.append(str(real_start+1)+".."+str(end))
result.append(str(L1_post))
result.append(Seq(str(L1_post)).translate())
else:
real_start=start+M.start()+M.start()+M.start()
result.append(type)
result.append(DEF.group(1))
result.append(str(real_start+1)+".."+str(end))
result.append(genomic_sequence[int(real_start):int(end)])
result.append(query)
return result
def do_blast(self, subject):
print 'Running BLAST...'
blast = blastp()
blast.db = self.standard
blast.query = subject
blast.outfmt = 5
blast.out = subject + '.xml'
blast.comp_based_stats = "0"
if os.path.exists(blast.out):
return
blast()
def secondary_blast(self):
if os.path.exists('orthologs.faa') is False or os.path.exists('secondary.xml') is True:
return
os.system('makeblastdb -dbtype prot -in orthologs.faa')
blast = blastp()
blast.db = 'orthologs.faa'
blast.query = self.substract
blast.outfmt = 5
blast.out = 'secondary.xml'
blast.evalue = self.expect_diff
blast.comp_based_stats = "0"
blast()
def primary_blast(self):
if os.path.exists('primary.xml') is True:
return
print 'Running BLAST...'
blast = blastp()
blast.db = self.db
blast.query = self.genomes[0][1]
blast.outfmt = 5
blast.out = 'primary.xml'
blast.evalue = self.expect
blast.comp_based_stats = "0"
blast()
def circle_blast(self, loc):
print 'Running Circualr BLAST...'
blast = blastp()
blast.db = loc + '/circle'
blast.outfmt = 5
blast.comp_based_stats = "0"
for subject in self.genomes:
self.build_genome(loc, subject)
blast.query = subject
blast.out = loc + '/' + subject + '.xml'
if os.path.exists(blast.out):
return
blast()
def match_fasta_position(query, subject, num=None):
"""return dataframe of position matching in subject and query fasta, or a dictionary given a list of number"""
df = []
blastp(query=query,
subject=subject,
out=query + '_' + subject + '.xml',
outfmt=5,
max_hsps=1)()
xml = SearchIO.read(query + '_' + subject + '.xml', "blast-xml")
for n in range(len(xml)):
x = xml[n][0]
hit_gap = np.array(
[index for index, value in enumerate(x.hit) if value == '-'])
q_gap = np.array(
[index for index, value in enumerate(x.query) if value == '-'])
hit_num = list(np.arange(x.hit_start + 1, x.hit_end + 1))
q_num = list(np.arange(x.query_start + 1, x.query_end + 1))
for i in hit_gap:
hit_num.insert(i, np.nan)
for i in q_gap:
q_num.insert(i, np.nan)
df.append(
list(
zip(
len(x.query) * [query],
len(x.query) * xml[n].description, x.query, q_num, x.hit,
hit_num)))
df = pd.DataFrame([j for i in df for j in i],
columns=[
'query', 'sequence', 'query_res', 'query_num',
'hit_res', 'hit_num'
])
df.set_index(['query'], inplace=True)
os.remove(query + '_' + subject + '.xml')
if num == None:
return df
else:
return df[df.query_num.isin(list(map(int, num)))].to_dict('records')
def run_blastp(query, subject, outfile, evalue=1e-4):
cmd = blastp(query=query,
subject=subject,
evalue=evalue,
outfmt=6,
out=outfile)
print(cmd)
stdout, stderr = cmd()
#if stderr:
# logging.warn("STDERR = ", stderr)
if not os.path.isfile(outfile):
raise Exception('No BLAST output')
if os.path.getsize(outfile) == 0:
return 0
return len(open(outfile).read().splitlines())
def blast_gene(seq, database):
tempfasta = open('temp.fasta', 'w')
SeqIO.write(seq, tempfasta, 'fasta')
tempfasta.close()
run = blastp(query='temp.fasta',
db=database,
max_target_seqs=1,
num_threads=6,
outfmt=5,
out='temp.xml')
run()
result_handle = open('temp.xml')
result = NCBIXML.read(result_handle)
if len(result.descriptions) > 0:
rets = result.descriptions[0].title
rets = [rets[rets.find('GQ'):]]
rets.append(str(result.descriptions[0].e))
return rets
else:
return ['none']
def blast_gene(ids, eval, database, of):
fasta_sequences = SeqIO.parse(open(database), "fasta")
for seq in fasta_sequences:
if seq.id == ids:
SeqIO.write(seq, "temp.fasta", "fasta")
run = blastp(query='temp.fasta',
db=database,
num_threads=6,
outfmt=5,
word_size=4,
evalue=eval,
out='temp.xml')
run()
result_handle = open('temp.xml')
result = NCBIXML.read(result_handle)
rets = []
for i in result.descriptions:
ttl = i.title
e = i.e
species = ttl.split(' ')[0]
rets.append(species)
rets.append(str(e))
# for i in result.alignments:
# for j in i.hsps:
# rets.append(str(j.frame[1]))
# rets.append(str(j.query))
# rets.append(str(j.match))
# rets.append(str(j.sbjct_start))
os.remove('temp.fasta')
os.remove('temp.xml')
genlis = []
for i in range(0, len(rets), 2):
genlis.append(rets[i])
print rets[i]
fasta_sequences = SeqIO.parse(open(database), "fasta")
seqs = []
for seq in fasta_sequences:
if seq.id in genlis:
seqs.append(seq)
SeqIO.write(seqs, of, "fasta")
def Blast(type, protein_sequence, start, end, genomic_sequence):
result = []
ORF = []
M = re.search('M', protein_sequence)
if M:
query = protein_sequence[M.start():]
query = query + "*"
temp = open("temp.ORF", "w")
print >> temp, '>blasting'
print >> temp, query
temp.close()
cline = blastp(query="'temp.ORF'",
db="./Blast/DB.blast.txt",
evalue=0.01,
outfmt=5,
out=type + ".BLAST")
os.system(str(cline))
blast_out = open(type + ".BLAST")
string = str(blast_out.read())
DEF = re.search("<Hit_def>((.*))</Hit_def>", string)
if DEF:
if DEF.group(1) == 'L1':
real_start = start + M.start() + M.start() + M.start()
result.append(type)
result.append('L1')
L1_pre = genomic_sequence[(start + 3 * M.start()):int(end)]
splice = '(C|T)(C|T)(A|C|G|T)(C|T)AG(A)TG'
spliced = re.search(splice, str(L1_pre))
if spliced:
start_L1 = int(spliced.start()) + 6
if start_L1 % 3 == 0:
if start_L1 > 600:
L1_post = L1_pre
result.append(str(start_L1) + ".." + str(end))
result.append(str(L1_post))
result.append(Seq(str(L1_post)).translate())
else:
L1_post = L1_pre[start_L1:]
result.append(
str(int(real_start + 1) + int(start_L1)) +
".." + str(end))
result.append(str(L1_post))
result.append(Seq(str(L1_post)).translate())
else:
L1_post = L1_pre
result.append(str(real_start + 1) + ".." + str(end))
result.append(str(L1_post))
result.append(Seq(str(L1_post)).translate())
else:
L1_post = L1_pre
result.append(str(real_start + 1) + ".." + str(end))
result.append(str(L1_post))
result.append(Seq(str(L1_post)).translate())
else:
real_start = start + M.start() + M.start() + M.start()
result.append(type)
result.append(DEF.group(1))
result.append(str(real_start + 1) + ".." + str(end))
result.append(genomic_sequence[int(real_start):int(end)])
result.append(query)
return result
def get_bsr(work_path,
job_name,
reference_fasta,
comparison_fastas,
strain_names,
force_redo=False,
matrix='PAM30'):
"""Return a BSR_RecordSet containing BSR_Record objects with
the results of a BSR analysis.
Args:
work_path (str): The path were generated files will be saved to.
job_name (str): Prefix used for all generated file names, and stored
in the returned BSR_Record
reference_fasta (str): Path of the FastA file used as the reference
comparison_fastas (iter[str]): List or tuple of paths to FastA files
for which BSR will be generated.
strain_names (iter[str]): Strain/species names of organisms in the
analysis. Used when generating graphs and optionally to obtain
data of that strain.
force_redo (bool): If True the analysis will be performed from the
start, overwriting any files in the workpath. Defaults to False.
matrix (str) optional, 'PAM30' (default) or 'BLOSUM62': Specifies the
scoring matrix used by BLAST. PAM30 is better for more similar
sequences.
Returns:
BSR_RecordSet. See documentation for this object for more info
A series of blastp are performed using the reference_fasta as the query
against the comparison_fastas. Files archiving the process are created
and saved to disc using pickle. If the process is interupted then it
can be continued by running the same command.
"""
assert len(comparison_fastas) == len(strain_names) - 1
#matrix = MATRICES[matrix]
created_files = []
# Create (or load) the initial BSR_records from the prime fasta
fprefix = work_path + job_name
init_bsr_records_fn = fprefix + ' - ref BSR records.pickle'
bsr_records = _do_next_step_pickle(
init_bsr_records_fn,
get_initial_BSR_Records_with_ref_scores,
(reference_fasta, strain_names[0],
matrix), # Don't forget to pass *args as tuple
force_redo=force_redo)
created_files.append(init_bsr_records_fn)
# Go through each proteome to be compared to the reference and get BSR.
if type(comparison_fastas) is not list:
comparison_fastas = [comparison_fastas]
comparison_num = 0
for comparee_index, subject_fasta in enumerate(comparison_fastas):
comparison_num += 1
# Get a blast of every reference sequences against the comparison proteome
blast_results_path = fprefix + ' - BLAST ref vs ' + os.path.split(
subject_fasta)[1] + '.xml'
created_files.append(blast_results_path)
if not os.path.isfile(blast_results_path) or force_redo:
blast_query = blastp(query=reference_fasta,
subject=subject_fasta,
remote=False,
outfmt=5,
max_hsps=1,
matrix=matrix,
comp_based_stats="0")
# run the blast
blast_results = blast_query()
with open(blast_results_path, 'w') as f:
f.write(blast_results[0])
print('BLAST results file saved:', blast_results_path)
bsr_records_path = fprefix + ' - {} of {} BSR calculated.pickle'.format(
comparison_num, len(comparison_fastas))
created_files.append(bsr_records_path)
bsr_records = _do_next_step_pickle(bsr_records_path,
get_bsr_for_strain,
(bsr_records, blast_results_path,
strain_names[comparee_index + 1]),
force_redo=force_redo)
final_results_path = fprefix + ' - complete BSR results.pickle'
created_files.append(final_results_path)
final_results = BSR_RecordSet(job_name, bsr_records, reference_fasta,
comparison_fastas, strain_names,
created_files, matrix)
fn = work_path + job_name + ' - Final record set.pickle'
with open(fn, 'wb') as f:
pickle.dump(final_results, f)
print('File saved:', fn)
return final_results