def load_ids(filename, filetype):
if filetype == "tabular":
for line in open(filename):
line = line.rstrip("\n")
if line and not line.startswith("#"):
yield line.split("\t", 1)[0]
elif filetype == "fasta":
for line in open(filename):
if line.startswith(">"):
yield line[1:].rstrip("\n").split(None, 1)[0]
elif filetype.startswith("fastq"):
# Use the Galaxy library not Biopython to cope with CS
from galaxy_utils.sequence.fastq import fastqReader
handle = open(filename, "rU")
for record in fastqReader(handle):
# The [1:] is because the fastaReader leaves the @ on the identifer.
yield record.identifier.split()[0][1:]
handle.close()
elif filetype == "sff":
try:
from Bio.SeqIO import index
except ImportError:
sys.exit("Require Biopython 1.54 or later (to read SFF files)")
# This will read the SFF index block if present (very fast)
for name in index(filename, "sff"):
yield name
else:
sys.exit("Unexpected file type %s" % filetype)
def load_ids(filename, filetype):
if filetype == "tabular":
for line in open(filename):
line = line.rstrip("\n")
if line and not line.startswith("#"):
yield line.split("\t", 1)[0]
elif filetype == "fasta":
for line in open(filename):
if line.startswith(">"):
yield line[1:].rstrip("\n").split(None, 1)[0]
elif filetype.startswith("fastq"):
# Use the Galaxy library not Biopython to cope with CS
from galaxy_utils.sequence.fastq import fastqReader
handle = open(filename, "rU")
for record in fastqReader(handle):
# The [1:] is because the fastaReader leaves the @ on the identifer.
yield record.identifier.split()[0][1:]
handle.close()
elif filetype == "sff":
try:
from Bio.SeqIO import index
except ImportError:
sys.exit("Require Biopython 1.54 or later (to read SFF files)")
# This will read the SFF index block if present (very fast)
for name in index(filename, "sff"):
yield name
else:
sys.exit("Unexpected file type %s" % filetype)
def main():
tree = read(sys.argv[1], 'newick')
seqs = index(sys.argv[2], 'fasta')
if not tree.rooted:
tree.root_at_midpoint()
tree.ladderize(reverse=True)
for leaf in tree.get_terminals():
write(seqs[leaf.name], sys.stdout, 'fasta')
hmmResultsDir = "hmm/results"
# crisprFiles = load(open("pickles/CRISPRs.p","rb"))
# casOperons = CasOperons(gene)
# casOperons.hasCas9(hmmResultsDir,crisprFiles)
casOperons = load(open("pickles/Cas9_Operons_HMM.p", 'rb'))
#Get unique chrs and the proteins they are associated with
allCasAsmFile = "assemblies/All_%s_Unique_Assemblies.fasta" % (gene)
allCasAAsFile = "proteins/All_%s-Like.faa" % (gene)
casOperons.uniqueNukeSeqs(allCasAsmFile,
allCasAAsFile) # Calls dump when it finishes
# Launch the domain search for the faa file created above
system(
"sbatch /mnt/research/germs/shane/transActRNA/scripts/hpc/DomainSearch.sb")
casAAs = dict(index(allCasAAsFile, "fasta"))
unUsed = set(casOperons.seqMap.protToAsm).difference(casAAs)
deletedOperons = {}
pres, absnt, hasSeq = 0, 0, 0
for protID in unUsed:
try:
operon = casOperons.operons[casOperons.seqMap[protID]]
deletedOperons[protID] = operon
except:
absnt += 1
from pickle import dump
dump(
deletedOperons,
open("/mnt/research/germs/shane/transActRNA/data/pickles/DeletedOperons.p",
"wb"))
return right_dist
else:
return -left_dist
if __name__ == "__main__":
data = pd.read_table(sys.argv[1],
sep='\s+',
comment='#',
names=[
'feature_id', '_1', '_2', '_3', 'position', '_5',
'_6', '_7', 'score', '_9', '_10', '_11'
],
index_col='feature_id',
usecols=['feature_id', 'position', 'score'])
seqindex = index(sys.argv[2], 'fasta')
for feature_id, (_position, score) in data.iterrows():
position = int(_position)
assert position == _position
position -= 1
seq = seqindex[feature_id].seq
c_pos = relative_pos_closest(str(seq), position, 'C')
if c_pos is None:
# Print an empty string if there is no 'C'.
c_pos = ''
print(
feature_id,
position,
round(score, 3),
type=int,
dest='min_length',
default=1000,
help="Minimum contig length to output [%(default)s]")
p.add_argument('seq_path',
type=str,
metavar="FASTA",
help="Sequences to be trimmed.")
p.add_argument(
'corr_handle',
type=argparse.FileType('r'),
metavar="CORR",
help="Correlation table from calculate_per_position_stats.py")
args = p.parse_args()
seqs = index(args.seq_path, 'fasta')
data = pd.read_table(
args.corr_handle,
names=['contig_id', 'position', 'total_depth', 'cosine_similarity'])
data.contig_id = data.contig_id.astype(str) # FIXME: Brittle
data.position = data.position - 1 # Convert to zero-indexed.
tally_seqs = 0
tally_nucs = 0
for contig_id in tqdm(list(seqs.keys())):
seq = seqs[contig_id].seq
if len(seq) < args.min_length:
# print(f"Contig {contig_id} too short.", file=sys.stderr)
continue
d = data[data.contig_id == contig_id]
if d.empty:
print("\rWARNING: {} not found in corr data.".format(contig_id),
#!/usr/bin/env python3
from Bio.SeqIO import index
from copy import deepcopy
import sys
if __name__ == "__main__":
seq_index = index(sys.argv[1], 'fasta')
for frag in sys.argv[2:]:
seq_id, *indices = frag.rsplit(':', 1)
rec = deepcopy(seq_index[seq_id])
if indices:
left, right = indices[0].split('-')
left = int(left)
right = int(right)
else:
left, right = 0, len(rec)
if left > right:
rec.seq = rec.seq[right:left].reverse_complement()
else:
rec.seq = rec.seq[left:right]
print(f'>{rec.id}\n{rec.seq}')
for codon in TRANSLATION:
POSITIONS[codon] = _count_positions(codon)
def count_positions(codon):
return POSITIONS[codon]
if __name__ == "__main__":
p = argparse.ArgumentParser()
p.add_argument('align1', metavar='FASTA1')
p.add_argument('align2', metavar='FASTA2', nargs='?')
args = p.parse_args()
if args.align2:
rec_index1 = index(args.align1, 'fasta')
rec_index2 = index(args.align2, 'fasta')
comparisons = []
for idA, idB in zip(rec_index1.keys(), rec_index2.keys()):
comparisons.append((idA, rec_index1[idA], idB, rec_index2[idB]))
else:
rec_index = index(args.align1, 'fasta')
comparisons = []
ids = list(rec_index.keys())
for i, idA in enumerate(ids):
for j, idB in enumerate(ids[i + 1:]):
comparisons.append((idA, rec_index[idA], idB, rec_index[idB]))
for idA, recA, idB, recB in comparisons:
seqA = recA.seq
seqB = recB.seq
gene = "Cas9"
assemblyDir = "assemblies/assemblies_W_%s/" % (gene)
cas9Assemblies = listdir(assemblyDir)
goodDomIDS = load(open("pickles/%s_GoodDomainIDS.p" % (gene), "rb"))
goodDomMap = load(open("pickles/%s_GoodDomMap.p" % (gene), "rb"))
hmm_parser = load(open("pickles/%s_HMM_Parsing_Results.p" % (gene), "rb"))
print("All loaded")
#Copy unique nucleotide sequence
from Bio.SeqIO import index
nukSeqHash, protSeqHash = set(), set()
alreadyGotIt, count = 0, 0
for assembly in cas9Assemblies:
baseID = assembly[:-6]
allAssemblySeqs = index(assemblyDir + assembly, "fasta")
overlap = goodDomIDS.intersection(allAssemblySeqs.keys())
for recID in overlap:
seq = str(allAssemblySeqs[recID].seq).upper()
if seq in nukSeqHash and len(goodDomMap[recID]) == 1:
alreadyGotIt += 1
continue
nukSeqHash.add(seq)
#There may be more than 1 protein on the pseudochromosome, save both as separate files
if len(goodDomMap[recID]) > 1:
print("%i Cas9s on %s %s" %
(len(goodDomMap[recID]), recID, baseID))
for orfID in goodDomMap[recID]:
# protSeq = str(hmm_parser.results[baseID].proteins[orfID].seq).upper()
with open("assemblies/pseudoChromos/%s.fasta" % (orfID),
"w") as fh:
