def get_rask_var(raskFasta, contig_file, fileName, outdir, verbose):
#run blast against the rask VAR genes
blast_out = run_blast(raskFasta, contig_file, outdir, verbose)
rask_hits = set()
with open(blast_out, 'r') as blastfile:
for line in blastfile:
rask_hits.add(line.split()[0])
rask = outdir + fileName + "_rask.fa"
non_rask = outdir + fileName + "_nonrask.fa"
count_rask = 0
count_non_rask = 0
with open(rask, 'w') as raskout:
with open(non_rask, 'w') as nonraskout:
for h, s in FastaReader(contig_file):
if h in rask_hits:
raskout.write(">" + h + "\n")
raskout.write(s + "\n")
count_rask += 1
else:
nonraskout.write(">" + h + "\n")
nonraskout.write(s + "\n")
count_non_rask += 1
if verbose:
print count_rask, " annotated to rask DB..."
print count_non_rask, " remaining"
return rask, non_rask
def analyse_contigs(contig_file,
read1,
read2,
outputdir,
fasta_ref_files=[],
verbose=False):
#first prepare blast files for analysis
renamedContigs = reNameContigs(contig_file, outputdir)
blast_files = []
for reference in fasta_ref_files:
blast_files.append(
run_blast(reference, renamedContigs, outputdir, verbose))
#now align reads to contigs
samfile = align_w_subread(read1, read2, renamedContigs, outputdir, verbose)
#now convert to bam
bamfile = convert_to_bam_create_index(renamedContigs, samfile, verbose)
# bamfile="notused"
#now compute analytics
outfile = generate_summary(bamfile, outputdir, renamedContigs, blast_files)
return outfile
def get_rask_var(raskFasta, contig_file, fileName, outdir, verbose):
#run blast against the rask VAR genes
blast_out = run_blast(raskFasta, contig_file, outdir, verbose)
rask_hits = set()
with open(blast_out , 'r') as blastfile:
for line in blastfile:
rask_hits.add(line.split()[0])
rask = outdir + fileName + "_rask.fa"
non_rask = outdir + fileName + "_nonrask.fa"
count_rask = 0
count_non_rask = 0
with open(rask, 'w') as raskout:
with open (non_rask, 'w') as nonraskout:
for h,s in FastaReader(contig_file):
if h in rask_hits:
raskout.write(">"+h+"\n")
raskout.write(s+"\n")
count_rask += 1
else:
nonraskout.write(">"+h+"\n")
nonraskout.write(s+"\n")
count_non_rask +=1
if verbose:
print count_rask, " annotated to rask DB..."
print count_non_rask, " remaining"
return rask, non_rask
def get_contaminants(fasta_ref_files, contig_file, fileName, percent_overlap
, outdir, verbose):
#first get list of contigs
contigs = {}
for h,s in FastaReader(contig_file):
contigs[h] = s
if verbose:
print ("Number of contigs before contaminant filtering: "
, len(contigs.keys()))
#now run blast against the reference files which we want not to be
#present in the data i.e. human
blast_files = []
for reference in fasta_ref_files:
blast_files.append(run_blast(reference, contig_file, outdir
, verbose))
#now iterate through blast results file removing contigs that have to
#high a proportion of hits
bad_contigs = set()
for blast_file in blast_files:
blast_name = os.path.splitext(os.path.basename(blast_file))[0]
with open(blast_file, 'r') as bfile:
for line in bfile:
tokens = line.strip().split()
name = tokens[0]
overlap = int(tokens[3])/float(len(contigs[name]))
if overlap > percent_overlap:
#we don't want this contig
print "removing", name, "overlapped", blast_name
bad_contigs.add(name)
#now write out a fasta file of contaminant sequences
contaminant_file = outdir + fileName + "contaminants.fa"
with open(contaminant_file, 'w') as outfile:
for contig in bad_contigs:
outfile.write(">" + contig + "\n")
outfile.write(contigs[contig] + "\n")
#now write contigs without contaminants to a file
non_contaminant_file = outdir + fileName + "Non_contaminants.fa"
with open(non_contaminant_file, 'w') as outfile:
for contig in contigs:
if contig not in bad_contigs:
outfile.write(">" + contig + "\n")
outfile.write(contigs[contig] + "\n")
if verbose:
print ("Number of contigs after filtering: "
, len(contigs.keys())-len(bad_contigs))
return non_contaminant_file, contaminant_file
def get_contaminants(fasta_ref_files, contig_file, fileName, percent_overlap,
outdir, verbose):
#first get list of contigs
contigs = {}
for h, s in FastaReader(contig_file):
contigs[h] = s
if verbose:
print("Number of contigs before contaminant filtering: ",
len(contigs.keys()))
#now run blast against the reference files which we want not to be
#present in the data i.e. human
blast_files = []
for reference in fasta_ref_files:
blast_files.append(run_blast(reference, contig_file, outdir, verbose))
#now iterate through blast results file removing contigs that have to
#high a proportion of hits
bad_contigs = set()
for blast_file in blast_files:
blast_name = os.path.splitext(os.path.basename(blast_file))[0]
with open(blast_file, 'r') as bfile:
for line in bfile:
tokens = line.strip().split()
name = tokens[0]
overlap = int(tokens[3]) / float(len(contigs[name]))
if overlap > percent_overlap:
#we don't want this contig
print "removing", name, "overlapped", blast_name
bad_contigs.add(name)
#now write out a fasta file of contaminant sequences
contaminant_file = outdir + fileName + "contaminants.fa"
with open(contaminant_file, 'w') as outfile:
for contig in bad_contigs:
outfile.write(">" + contig + "\n")
outfile.write(contigs[contig] + "\n")
#now write contigs without contaminants to a file
non_contaminant_file = outdir + fileName + "Non_contaminants.fa"
with open(non_contaminant_file, 'w') as outfile:
for contig in contigs:
if contig not in bad_contigs:
outfile.write(">" + contig + "\n")
outfile.write(contigs[contig] + "\n")
if verbose:
print("Number of contigs after filtering: ",
len(contigs.keys()) - len(bad_contigs))
return non_contaminant_file, contaminant_file
def filter_ref_with_blast(fasta_ref_files, contig_file, percent_overlap
, outfile, outdir):
#first get list of contigs
contigs = {}
for h,s in FastaReader(contig_file):
contigs[h] = s
print "Number of contigs before filtering: ", len(contigs.keys())
#now run blast against the reference files which we want not to be
#present in the data i.e. human
blast_files = []
for reference in fasta_ref_files:
blast_files.append(run_blast(reference, contig_file, outdir
, True))
#now iterate through blast results file removing contigs that have to
#high a proportion of hits
bad_contigs = set()
for blast_file in blast_files:
blast_name = os.path.splitext(os.path.basename(blast_file))[0]
with open(blast_file, 'r') as bfile:
for line in bfile:
tokens = line.strip().split()
name = tokens[0]
overlap = int(tokens[3])/float(len(contigs[name]))
if overlap > percent_overlap:
#we don't want this contig
print "removing", name, "overlapped", blast_name
bad_contigs.add(name)
for name in bad_contigs:
del contigs[name]
#now write resulting contigs to a file
with open(outfile, 'w') as outfas:
for contig in contigs:
outfas.write(">" + contig + "\n")
outfas.write(contigs[contig] + "\n")
print "Number of contigs after filtering: ", len(contigs.keys())
def filter_ref_with_blast(fasta_ref_files, contig_file, percent_overlap,
outfile, outdir):
#first get list of contigs
contigs = {}
for h, s in FastaReader(contig_file):
contigs[h] = s
print "Number of contigs before filtering: ", len(contigs.keys())
#now run blast against the reference files which we want not to be
#present in the data i.e. human
blast_files = []
for reference in fasta_ref_files:
blast_files.append(run_blast(reference, contig_file, outdir, True))
#now iterate through blast results file removing contigs that have to
#high a proportion of hits
bad_contigs = set()
for blast_file in blast_files:
blast_name = os.path.splitext(os.path.basename(blast_file))[0]
with open(blast_file, 'r') as bfile:
for line in bfile:
tokens = line.strip().split()
name = tokens[0]
overlap = int(tokens[3]) / float(len(contigs[name]))
if overlap > percent_overlap:
#we don't want this contig
print "removing", name, "overlapped", blast_name
bad_contigs.add(name)
for name in bad_contigs:
del contigs[name]
#now write resulting contigs to a file
with open(outfile, 'w') as outfas:
for contig in contigs:
outfas.write(">" + contig + "\n")
outfas.write(contigs[contig] + "\n")
print "Number of contigs after filtering: ", len(contigs.keys())
def analyse_contigs(contig_file, read1, read2, outputdir
, fasta_ref_files=[], verbose=False):
#first prepare blast files for analysis
renamedContigs = reNameContigs(contig_file, outputdir)
blast_files = []
for reference in fasta_ref_files:
blast_files.append(run_blast(reference, renamedContigs, outputdir
, verbose))
#now align reads to contigs
samfile = align_w_subread(read1, read2, renamedContigs, outputdir, verbose)
#now convert to bam
bamfile = convert_to_bam_create_index(renamedContigs, samfile, verbose)
# bamfile="notused"
#now compute analytics
outfile = generate_summary(bamfile, outputdir
, renamedContigs, blast_files)
return outfile
import sys, os
from third_party_runners import run_blast
from mungo.fasta import FastaReader
OVERLAP = 0.7
IDENTITY = 95
sequences = sys.argv[1]
outdir = sys.argv[2]
#first run all-vs-all blast
bfile = run_blast(sequences, sequences, outdir, True)
#now get the lengths of all the contigs
contig_len = {}
for h,s in FastaReader(sequences):
contig_len[h]=len(s)
#now iterate through search finding redundant contigs
redundant_contigs = set()
with open(bfile, 'r') as blastsearch:
for line in blastsearch:
tokens = line.strip().split()
if tokens[0]==tokens[1]: #matching itself
continue
if float(tokens[2]) >= IDENTITY:
# print "identity ",tokens[2]
if ((float(tokens[3])/contig_len[tokens[0]] >= OVERLAP) or
(float(tokens[3])/contig_len[tokens[1]] >= OVERLAP)):
# print "overlap", max(float(tokens[3])/contig_len[tokens[0]],float(tokens[3])/contig_len[tokens[1]])
import sys, os
from third_party_runners import run_blast
from mungo.fasta import FastaReader
contig_file = sys.argv[1]
blastdb_file = sys.argv[2]
outdir = sys.argv[3]
name = os.path.splitext(os.path.basename(contig_file))[0]
blast_out = run_blast(blastdb_file, contig_file, outdir, True)
rask_hits = set()
with open(blast_out , 'r') as blastfile:
for line in blastfile:
rask_hits.add(line.split()[0])
with open(outdir + name + "_rask.fa", 'w') as raskout:
with open (outdir + name + "_nonrask.fa", 'w') as nonraskout:
for h,s in FastaReader(contig_file):
if h in rask_hits:
raskout.write(">"+h+"\n")
raskout.write(s+"\n")
else:
nonraskout.write(">"+h+"\n")
nonraskout.write(s+"\n")
import sys, os
from third_party_runners import run_blast
from mungo.fasta import FastaReader
contig_file = sys.argv[1]
blastdb_file = sys.argv[2]
outdir = sys.argv[3]
name = os.path.splitext(os.path.basename(contig_file))[0]
blast_out = run_blast(blastdb_file, contig_file, outdir, True)
rask_hits = set()
with open(blast_out, 'r') as blastfile:
for line in blastfile:
rask_hits.add(line.split()[0])
with open(outdir + name + "_rask.fa", 'w') as raskout:
with open(outdir + name + "_nonrask.fa", 'w') as nonraskout:
for h, s in FastaReader(contig_file):
if h in rask_hits:
raskout.write(">" + h + "\n")
raskout.write(s + "\n")
else:
nonraskout.write(">" + h + "\n")
nonraskout.write(s + "\n")