print "failed to handle Genbank file"
break
else:
print "...",
seq_format = 'gbk'
elif filename.find(".fas") > 0:
# process fasta (for mfas, load first record)
try:
record = load_fasta(seq_dir+filename)
except IOError:
print "failed to load Fasta file as single-record file"
break
except Exception:
try:
record = load_multifasta(seq_dir+filename)[0]
except IOError:
print "failed to load Fasta file as multi-record file"
break
except Exception:
print "failed to handle Fasta file"
break
print "...",
seq_format = 'fas'
else:
# reject as bad format
print "invalid file format"
break
if len(record) < int(min_size):
# script to translate sequences in multifasta files into proteins
from sys import argv
from libs.common import load_multifasta, write_fasta
from Bio.SeqRecord import SeqRecord
origin_dir = "data/" + argv[1] + "/"
in_file = origin_dir + argv[2]
outfile = in_file[:-4] + "_aa.fas"
proteins = []
for record in load_multifasta(in_file):
aa_rec = SeqRecord(id=record.id, seq=record.seq.translate())
proteins.append(aa_rec)
write_fasta(outfile, proteins)
data_dir = "data/"+argv[1]+"/"
dir_in = data_dir+argv[2]+"/"
infile = data_dir+argv[3] # must be a fasta file with query sequences
file_ext = argv[4]
blast_mode = argv[5]
if len(argv) > 5:
blast_mode = argv[5]
else:
blast_mode = 'n' # nucleotide blast by default
blast_out = data_dir+"blast_out/"
ensure_dir([blast_out])
queries = load_multifasta(infile)
filenames = from_dir(dir_in, re.compile(r'.*\.'+file_ext))
for filename in filenames:
rec_name = filename[:filename.find("."+file_ext)]
print rec_name,
genome_path = dir_in+filename
dbfile_path = "data/blast_db/"+rec_name
while True:
if not path.exists(dbfile_path+".nhr"):
if file_ext == 'gbk':
try:
# script to rename contigs in multifasta files
<<<<<<< HEAD
from genomes import all as genomes
from libs.common import load_multifasta, write_fasta
for genome in genomes:
print genome['file']
file_path = "data/genomes/"+genome['file']
outfile_path = "data/renamed/"+genome['file']
contigs = load_multifasta(file_path)
renamed = []
counter = 1
for contig in contigs:
contig.id = genome['name']+"_"+str(counter)
contig_path = "data/contigs/"+contig.id+".fas"
write_fasta(contig_path, contig)
renamed.append(contig)
counter +=1
write_fasta(outfile_path, renamed)
=======
from sys import argv
from libs.common import load_multifasta, write_fasta, ensure_dir
from genomes import all as genome_list
origin_dir = "data/"+argv[1]+"/"
destin_dir = "data/"+argv[2]+"/"
ensure_dir([destin_dir])
data_dir = "data/" + argv[1] + "/"
dir_in = data_dir + argv[2] + "/"
infile = data_dir + argv[3] # must be a fasta file with query sequences
file_ext = argv[4]
blast_mode = argv[5]
if len(argv) > 5:
blast_mode = argv[5]
else:
blast_mode = 'n' # nucleotide blast by default
blast_out = data_dir + "blast_out/"
ensure_dir([blast_out])
queries = load_multifasta(infile)
filenames = from_dir(dir_in, re.compile(r'.*\.' + file_ext))
for filename in filenames:
rec_name = filename[:filename.find("." + file_ext)]
print rec_name,
genome_path = dir_in + filename
dbfile_path = "data/blast_db/" + rec_name
while True:
if not path.exists(dbfile_path + ".nhr"):
if file_ext == 'gbk':
try:
import re
from sys import argv
from libs.common import load_multifasta, from_dir
import matplotlib.pyplot as plt
import numpy as np
data_dir = "data/" + argv[1]
filenames = from_dir(data_dir, re.compile(r'.*\.fas.*'))
ctg_ns = []
n50s = []
for filename in filenames:
# load contigs from file
contig_list = load_multifasta(data_dir + "/" + filename)
# count contigs
ctg_count = len(contig_list)
if ctg_count < 200:
ctg_ns.append(ctg_count)
else:
ctg_ns.append(200)
# sort contig list by size
contig_list.sort(key=len)
contig_list.reverse()
# count full sequence length
full_seq_length = 0
for contig in contig_list:
full_seq_length += len(contig.seq)
print "failed to handle Genbank file"
break
else:
print "...",
seq_format = 'gbk'
elif filename.find(".fas") > 0:
# process fasta (for mfas, load first record)
try:
record = load_fasta(seq_dir + filename)
except IOError:
print "failed to load Fasta file as single-record file"
break
except Exception:
try:
record = load_multifasta(seq_dir + filename)[0]
except IOError:
print "failed to load Fasta file as multi-record file"
break
except Exception:
print "failed to handle Fasta file"
break
print "...",
seq_format = 'fas'
else:
# reject as bad format
print "invalid file format"
break
if len(record) < int(min_size):
# script to translate sequences in multifasta files into proteins
from sys import argv
from libs.common import load_multifasta, write_fasta
from Bio.SeqRecord import SeqRecord
origin_dir = "data/"+argv[1]+"/"
in_file = origin_dir+argv[2]
outfile = in_file[:-4]+"_aa.fas"
proteins = []
for record in load_multifasta(in_file):
aa_rec = SeqRecord(id=record.id, seq=record.seq.translate())
proteins.append(aa_rec)
write_fasta(outfile, proteins)
else:
gbk_file = origin_dir+"/"+filename
fas_file = gbk2fas(gbk_file)
record = load_genbank(gbk_file)
# run prediction
annot_aa = annot_aa_dir+rec_name+"_ann.fas"
annot_gbk = annot_gbk_dir+rec_name+"_ann.gbk"
if not path.exists(trn_file):
train_prodigal(fas_file, trn_file, "-q")
if not path.exists(annot_aa):
run_prodigal(fas_file, annot_gbk, annot_aa, trn_file, "-q")
# collect orfs
record.features = []
aa_record = load_multifasta(annot_aa)
counter = 1
for aa_rec in aa_record:
this_prot = rec_name+"_"+str(counter)
# get feature details from description line
# because prodigal output fails to load as valid genbank
defline = aa_rec.description
pattern = re.compile('.+#\s(\d+)\s#\s(\d+)\s#\s(\S*1)\s#\sID.+')
match = pattern.match(defline)
start_pos = int(match.group(1))
end_pos = int(match.group(2))
strand_pos = int(match.group(3))
feat_loc = FeatureLocation(start_pos, end_pos)
l_tag = rec_name+"_"+str(counter)
# consolidation feature annotations
quals = {'note': defline, 'locus_tag': l_tag,
try:
records = [load_genbank(origin_file)]
except IOError:
print "failed to load file"
break
elif genome['input'] == 'fas':
try:
records = [load_fasta(origin_file)]
except IOError:
print "failed to load file"
break
elif genome['input'] == 'mfas':
try:
records = load_multifasta(origin_file)
except IOError:
print "failed to load file"
break
else:
print "input not recognized"
break
for record in records:
try:
write_fasta(destin_dir+record.id+".fas", record)
except Exception:
print "failed to write contig file"
break
else:
import re
from sys import argv
from libs.common import load_multifasta, from_dir
import matplotlib.pyplot as plt
import numpy as np
data_dir = "data/"+argv[1]
filenames = from_dir(data_dir, re.compile(r'.*\.fas.*'))
ctg_ns = []
n50s = []
for filename in filenames:
# load contigs from file
contig_list = load_multifasta(data_dir+"/"+filename)
# count contigs
ctg_count = len(contig_list)
if ctg_count < 200:
ctg_ns.append(ctg_count)
else:
ctg_ns.append(200)
# sort contig list by size
contig_list.sort(key=len)
contig_list.reverse()
# count full sequence length
full_seq_length = 0
for contig in contig_list:
full_seq_length += len(contig.seq)
record = load_genbank(gbk_file)
assert record.id
# run prediction
annot_aa = annot_aa_dir+rec_name+"_ann.fas"
annot_gbk = annot_gbk_dir+rec_name+"_ann.gbk"
if not path.exists(trn_file):
train_prodigal(fas_file, trn_file, "-q")
if not path.exists(annot_aa):
run_prodigal(fas_file, annot_gbk, annot_aa, trn_file, "-q")
# blast the protein sequences against the remote DB
record.features = []
evalue = 0.01
proteins = load_multifasta(annot_aa)
for protein in proteins:
print " ", protein.id
rec_hits_dir = hits_dir+rec_name+"/"
ensure_dir([rec_hits_dir])
hits_out = open(rec_hits_dir+protein.id+".txt", 'w')
hits_out.write(" ".join([protein.id, "vs.", remote_prot_db,
"@evalue =", str(evalue), "\n"]))
temp_out = remote_blastp_2file(protein.seq, remote_prot_db,
blast_dir+rec_name+"_temp.xml",
evalue)
#temp_out = blast_dir+rec_name+"_temp.xml"
# collect best 10 hits
rec_hits = collect_topNhits(temp_out, 10)
for hit in rec_hits:
if hasattr(hit, 'hsps'):
record = load_genbank(gbk_file)
assert record.id
# run prediction
annot_aa = annot_aa_dir + rec_name + "_ann.fas"
annot_gbk = annot_gbk_dir + rec_name + "_ann.gbk"
if not path.exists(trn_file):
train_prodigal(fas_file, trn_file, "-q")
if not path.exists(annot_aa):
run_prodigal(fas_file, annot_gbk, annot_aa, trn_file, "-q")
# blast the protein sequences against the remote DB
record.features = []
evalue = 0.01
proteins = load_multifasta(annot_aa)
for protein in proteins:
print " ", protein.id
rec_hits_dir = hits_dir + rec_name + "/"
ensure_dir([rec_hits_dir])
hits_out = open(rec_hits_dir + protein.id + ".txt", 'w')
hits_out.write(" ".join([
protein.id, "vs.", remote_prot_db, "@evalue =",
str(evalue), "\n"
]))
temp_out = remote_blastp_2file(protein.seq, remote_prot_db,
blast_dir + rec_name + "_temp.xml",
evalue)
#temp_out = blast_dir+rec_name+"_temp.xml"
# collect best 10 hits
rec_hits = collect_topNhits(temp_out, 10)