def detect_mutations(input: str, db: str, out: str):
os.system(
f"blastx -query {input} -subject {db} -subject_besthit -max_hsps 1 -outfmt 5 > tmp.blast.xml"
)
with open("tmp.blast.xml", "r") as f:
with open(out, "w") as fh:
done = 0
for p in parse(f):
mutations = []
for alignment in p.alignments:
al_mutations = []
hsps = alignment.hsps[0]
query = hsps.query
subject = hsps.sbjct
if (len(query)) != len(subject):
continue
for i in range(len(subject)):
if query[i] not in ['X', '-', '*'
] and query[i] != subject[i]:
al_mutations.append(
f"{alignment.title.split('|')[0]} {subject[i]}{i + hsps.sbjct_start}{query[i]}"
)
mutations.append(al_mutations)
done += 1
fh.write(f"{p.query};")
fh.write(','.join(
list(itertools.chain.from_iterable(mutations))))
fh.write("\n")
os.unlink("tmp.blast.xml")
def create_blast_dict():
#seq = acc2sequence(acc)
with open(blast_cache, mode="r") as f:
records = [record for record in parse(f)]
for record in records:
homologs = [{
"accession":
alignment.accession,
"identity":
round((hsp.identities / hsp.align_length) * 100, 2),
"coverage":
round((hsp.align_length / record.query_length) * 100, 2)
} for alignment in record.alignments for hsp in alignment.hsps]
with open(dict_cache, mode="wb") as f:
pickle.dump(homologs, f)
print("cached the homologs dictionary")
def calc_immutability(consensus_filename, protein_seqs_filename, CUTOFF):
out = NcbitblastnCommandline(subject=consensus_filename,
query=protein_seqs_filename,
out="results.xml",
outfmt=5) # evalue=?
out(
) # NCBI tblastn is run locally to align proteins to the consensus sequence
consensus_seq = SeqIO.parse(consensus_filename,
"fasta").__next__().seq._data
protein_alignments = [
x.alignments[0].hsps[0] for x in parse(open(out.out, "r"))
]
mutability_score = [UNCONSERVED_SCORE for _ in range(len(consensus_seq))]
for alignment in protein_alignments:
position = alignment.sbjct_start - 1
# print("\ncalculating for",alignment.sbjct)
for aa in alignment.sbjct:
assert aa != " ", "empty aa in alignment subject"
if aa != "X":
codon0 = consensus_seq[position:position + 3]
# print(aa, end="")
aa0 = "none"
try:
aa0 = codon_table[codon0]
except:
position += 3
continue #We do not calculate if there is an ambiguous nucleotide in the consensus seq. We could
#have calculated for the other nucleotides in the codon however it requires more
#detailed calculation.
# print(position,codon0,aa0,end="|")
for pos_in_codon in range(3):
# codonX=codon0
if mutability_score[position +
pos_in_codon] != UNCONSERVED_SCORE:
continue # if this position was previously processed, there is no reason to process again. SKIP IT. This might be the case if there are several proteins aligned to here.
mutability_score[position + pos_in_codon] = 0
for putative_nuc in SUBSTITUTION_MODEL[
codon0[pos_in_codon]].keys():
codonX = replace(codon0, putative_nuc, pos_in_codon)
aaX = codon_table[codonX]
if aaX != "_": #and aa0!=aaX: #To include self conversion sounds rigth way as both matrices include these in their calculations. Asuming any deleterious mutation would not be conserved thus not needed for consideration.
try:
aa_score = aa_score_method(aa0, aaX)
except:
aa_score = aa_score_method(
aaX, aa0
) #pam30 is a triangular matrix. so we check reversed
subs_score = SUBSTITUTION_MODEL[
codon0[pos_in_codon]][putative_nuc]
# print(aa0,"-->",aaX,":",aa_score*subs_score,end="|")
mutability_score[
position +
pos_in_codon] += aa_score * subs_score
position += 3
print("\nconsevation scores complete:")
standart_dev = std([x for x in mutability_score if x != None])
average = mean([x for x in mutability_score if x != None])
printlines = ["", "", "", "", "", ""]
for i in range(len(consensus_seq)):
if i % 100 == 0:
print(i, end=" " * (100 - len(str(i))))
printlines[0] += str(i) + " " * (100 - len(str(i)))
print()
for i in range(len(consensus_seq)):
if i % 100 == 0:
print("|", end="")
printlines[1] += "|"
else:
print(" ", end="")
printlines[1] += " "
print()
prot = "" #TODO: automatically find where protein starts, add all one by one. This code below assumes all has ORF +1.It would not work for all cases
for i in range(2, len(consensus_seq), 3):
d = " "
if mutability_score[i] != None:
try:
d = " " + codon_table[consensus_seq[i:i + 3]] + " "
except:
pass
prot += d
print(" " + prot)
# printlines[2]=" "+prot
for i in range(len(consensus_seq)):
print(consensus_seq[i], end="")
printlines[3] += consensus_seq[i]
print()
for i in range(len(consensus_seq)):
d = "_"
if mutability_score[i] != None:
d = " "
if mutability_score[i] < average + standart_dev * 2:
d = "░"
if mutability_score[i] < average + standart_dev:
d = "▒"
if mutability_score[i] < average - standart_dev:
d = "▓"
if mutability_score[i] < average - standart_dev * 2:
d = "█"
print(d, end="")
printlines[4] += d
print()
immutable_seq = ""
for i in range(len(consensus_seq)):
d = "N"
if mutability_score[i] != None and mutability_score[i] < CUTOFF:
d = consensus_seq[i]
print(d, end="")
printlines[5] += d
immutable_seq += d
# print()
# for i in range(0,30000,100):
# print("\n".join(printlines[i:i+100]))
print("\n<<<REPORT>>>")
print("settings - Score for unconserved nucleotides:", UNCONSERVED_SCORE,
"Cutoff score:", CUTOFF)
print("max score:", max([x for x in mutability_score if x != None]),
"min score:", min([x for x in mutability_score if x != None]))
print("mean score", average, "standart dev:", standart_dev)
print("number of non-Ns in consensus seq:",
len([x for x in consensus_seq if x != "N"]))
print("number of non-Ns in immutable seq:",
len([x for x in immutable_seq if x != "N"]))
print("scores of coding regions summed:",
sum([x for x in mutability_score if x != None]))
# print("scores of all regions summed:", sum([x for x in mutability_score]))
return immutable_seq, mutability_score
from Bio.Blast.NCBIWWW import qblast
from Bio.Blast.NCBIXML import parse
from Bio import SeqIO
records = SeqIO.parse("./apoe.fas", "fasta")
PROGRAM = 'blastp'
DATABASE = 'nr'
for rec in records:
# query NCBI Blast API
xml_result = qblast(PROGRAM, DATABASE, rec.seq)
# Parse xml result
results = parse(xml_result)
# iterate over each result
for record in results:
for alignment in record.alignments:
print(alignment)
def _main(self):
email = '*****@*****.**'
genome_dir = '/home/allis/Dropbox/Science/Микра/Thermococcus/sequence/GenBank/Thermococcales/Thermococcus/'
genome = 'Thermococcus_barophilus_Ch5.gb'
gene = 'TBCH5v1_1369' #cooS
database = 'nr'
segment = [3200, 12000]
seq = SeqLoader.load_file(os.path.join(genome_dir, genome))
if not seq: raise RuntimeError('No genome loaded')
seq = seq[0]
index = get_indexes_of_genes(seq, gene)
if not index: raise RuntimeError('No gene found')
feature = seq.features[index[0]]
query = feature.extract(seq)
segments_file = 'CO-clusters.gb'
#get cluster variants if needed
if not os.path.isfile(segments_file):
blast_file = 'blast.results.xml'
if os.path.isfile(blast_file):
blast = list(parse(open(blast_file)))
else:
blast = BlastCLI.blast_seq(query,
database,
100,
remote=True,
task='blastn',
parse_results=True,
save_results_to='blast.results.xml')
if not blast: raise RuntimeError('Blast returned no results')
flt = BlastFilter(lambda hsp, r: hsp.align_length > 700,
filter_hsps=True)
flt(blast)
queries = []
for ali in BlastCLI.iter_alignments(blast):
q = BlastCLI.Query(ali,
'hsp',
start_offset=segment[0],
end_offset=segment[1])
if q: queries.append(q)
print(queries[-1])
segments = BlastWWW.fetch_queries(email, queries)
safe_write(segments, segments_file)
for r in segments:
print('[%s] %s: %dbp' % (r.id, pretty_rec_name(r), len(r)))
return 0
#find primers in alignments of the selected features
local_files = [
os.path.join(genome_dir, f)
for f in ('Thermococcus_barophilus_DT4-complete-genome.gb',
'Thermococcus_ST-423.gb', 'Thermococcus_CH1-complete.gb')
]
loader = SeqLoader(self.abort_event)
segments = loader.load_files([segments_file] + local_files)
fprimers, transF_ali = find_primers(
segments, 'transF',
dict(plen=(20, 30),
max_mismatches=5,
min_first_matches=3,
AT_first=True))
rprimers, cooS_ali = find_primers(segments,
'cooS',
dict(plen=(20, 30),
max_mismatches=4,
min_first_matches=3,
AT_first=True),
reverse=True)
if not fprimers:
print('\nNo forward primers found')
return 1
if not rprimers:
print('\nNo reverse primers found')
return 1
print('\nForward primers:')
for p in fprimers:
print('%s: %s' % (p.id, p))
print('\nReverse primers:')
for p in rprimers:
print('%s: %s' % (p.id, p))
print()
#add primers to alignments and save them
transF_ali = PrimerFinder.add_primers_to_alignment(
fprimers, transF_ali)
cooS_ali = PrimerFinder.add_primers_to_alignment(rprimers,
cooS_ali,
reverse=True)
AlignmentUtils.save(transF_ali, 'transF.aln')
AlignmentUtils.save(cooS_ali, 'cooS.aln')