def split_seq_file(fa):
from os import makedirs, path
from fasta import fasta_iter
import gzip
makedirs('partials', exist_ok=True)
basename = path.basename(fa)
cur_n = MAX_SEQ_CHUNKS + 1
ix = 0
out = None
partials = []
for h, seq in fasta_iter(fa):
if cur_n >= MAX_SEQ_CHUNKS:
if fa.endswith('.faa.gz'):
partials.append(f'partials/{basename}_block_{ix:04}.faa.gz')
elif fa.endswith('.fna.gz'):
partials.append(f'partials/{basename}_block_{ix:04}.fna.gz')
else:
raise ValueError(f'Unexpected file name: {fa}')
out = gzip.open(partials[-1], compresslevel = 1, mode = 'wt')
cur_n = 0
ix += 1
out.write(f'>{h}\n{seq}\n')
cur_n +=1
out.close()
return partials
def dedup_fasta(infile):
from fasta import fasta_iter
import gzip
import os
print("start dedup")
fasta = {}
for ID, seq in fasta_iter(infile):
if seq in fasta:
fasta[seq][1] += 1
else:
fasta[seq] = [ID, 1]
outfile1 = infile.replace('.faa.gz', '.raw_number.tsv.gz')
outfile2 = infile.replace('.faa.gz', '.dedup.faa.gz')
out1 = gzip.open(outfile1, "wt", compresslevel=1)
out2 = gzip.open(outfile2, "wt", compresslevel=1)
print("start sort")
for seq, (ID, count) in sorted(fasta.items()):
out1.write(f"{count}\t{seq}\n")
out2.write(f">{ID}\n{seq}\n")
out1.close()
out2.close()
os.unlink(infile)
print("finish dedup and sort")
return (outfile1, outfile2)
def features(ifile):
groups = [set(g) for g in (GROUPS_SA + GROUPS_HB)]
seqs = []
headers = []
encodings = []
aaComp = []
for h, seq in fasta_iter(ifile):
if seq[-1] == '*':
seq = seq[:-1]
seqs.append(seq)
headers.append(h)
encodings.append(ctdd(seq, groups))
aaComp.append(amino_acid_composition(seq))
# We can do this inside the loop so that we are not forced to pre-load all
# the sequences into memory. However, it becomes much slower
rpy2.robjects.globalenv['seq'] = seqs
rfeatures = r('''
ch <- charge(seq=seq, pH=7, pKscale="EMBOSS")
pI <- pI(seq=seq, pKscale="EMBOSS")
aIndex <- aIndex(seq=seq)
instaIndex <- instaIndex(seq=seq)
boman <- boman(seq=seq)
hydrophobicity <- hydrophobicity(seq=seq, scale="Eisenberg")
hmoment <- hmoment(seq=seq, angle=100, window=11)
cbind(ch, pI, aIndex, instaIndex, boman, hydrophobicity, hmoment)
''')
features = np.hstack([aaComp, rfeatures, encodings])
features = pd.DataFrame(features,
index=headers,
columns=[
"tinyAA",
"smallAA",
"aliphaticAA",
"aromaticAA",
"nonpolarAA",
"polarAA",
"chargedAA",
"basicAA",
"acidicAA",
"charge",
"pI",
"aindex",
"instaindex",
"boman",
"hydrophobicity",
"hmoment",
"SA.Group1.residue0",
"SA.Group2.residue0",
"SA.Group3.residue0",
"HB.Group1.residue0",
"HB.Group2.residue0",
"HB.Group3.residue0",
])
features.insert(0, 'group', 'Unk')
features.insert(0, 'sequence', seqs)
return features
def mergeseq(outfile):
import heapq
import gzip
from glob import glob
from fasta import fasta_iter
with gzip.open(outfile, compresslevel=1, mode='wt') as output:
inputs = [fasta_iter(f) for f in glob(f'*.dedup.faa.gz')]
merged = heapq.merge(*inputs, key=lambda h_seq: (h_seq[1], h_seq[0]))
preseq = "x"
for h, seq in merged:
if seq != preseq:
output.write(f'>{h}\n{seq}\n')
preseq = seq
def read_fasta_sequences(filename):
"""
Function to parse text from the given FASTA file into a screed database
"""
# Will raise an exception if the file doesn't exist
theFile = open(filename, "rb")
# Setup the iterator function
iterfunc = fasta.fasta_iter(theFile)
# Create the screed db
create_db(filename, fasta.FieldTypes, iterfunc)
theFile.close()
return ScreedDB(filename)
def splitseq(infile):
print("start splitseq")
outputlist = [f'split/submetag_{ix:03}.faa.gz' for ix in range(256)]
outputfiles = [
gzip.open(f'split/submetag_{ix:03}.faa.gz', compresslevel=1, mode='wt')
for ix in range(256)
]
for ID, seq in fasta_iter(infile):
h = hashlib.sha256()
h.update(seq.encode('ascii'))
ix = int(h.hexdigest()[:2], 16)
outputfiles[ix].write(f'>{ID}\n{seq}\n')
for ot in outputfiles:
ot.close()
print("finish splitseq")
return (outputlist)
def calseq(infile, outfile1, outfile2):
from fasta import fasta_iter
import gzip
fasta = {}
for ID, seq in fasta_iter(infile):
if seq in fasta:
fasta[seq][1] += 1
else:
fasta[seq] = [ID, 1]
out1 = gzip.open(outfile1, "wt", compresslevel=1)
out2 = gzip.open(outfile2, "wt", compresslevel=1)
for seq, (ID, count) in fasta.items():
out1.write(f"{count}\t{seq}\n")
out2.write(f">{ID}\n{seq}\n")
out1.close()
out2.close()
def extract_seq(infile1,infile2,outfile1,outfile2):
from fasta import fasta_iter
import gzip
fastaset=set()
with gzip.open(infile1,"rt") as f:
for line in f:
line = line.strip()
linelist = line.split("\t")
if linelist[0] != "1":
fastaset.add(linelist[1])
with gzip.open(outfile1, "wt", compresslevel=1) as out1, \
gzip.open(outfile2, "wt", compresslevel=1) as out2:
for ID,seq in fasta_iter(infile2):
if seq in fastaset:
out1.write(f'>{ID}\n{seq}\n')
else:
out2.write(f'>{ID}\n{seq}\n')
def splitseq(infile, X, outfile):
'''split inputfile according to max number of sequences X'''
import gzip
from fasta import fasta_iter
ix = 0
n = 0
oname = outfile.format(ix=ix)
out = gzip.open(oname, "wt", compresslevel=1)
for ID, seq in fasta_iter(infile):
if n < X:
n += 1
out.write(f'>{ID}\n{seq}\n')
else:
n = 1
ix += 1
oname = outfile.format(ix=ix)
out = gzip.open(oname, "wt", compresslevel=1)
out.write(f'>{ID}\n{seq}\n')
if not ID:
break
out.close()