def main():
if len(sys.argv) < 4:
print >> sys.stderr, \
'Usage: assemgraph2.py <query file> <target file>' + \
' <blast9 file> [min score=200]'
raise SystemExit
try:
MIN_SCORE = float(sys.argv[5])
except IndexError:
pass
print >> sys.stderr, 'Reading sequence databases...'
queries = seqdb.SequenceFileDB(sys.argv[1])
targets = seqdb.SequenceFileDB(sys.argv[2])
print >> sys.stderr, len(queries), len(targets)
try:
align_file = open(sys.argv[3])
except IOError as e:
print >> sys.stderr, 'Error: check alignment file.'
raise e
print >> sys.stderr, 'Constructing alignment graphs...'
graph = nx.Graph()
for c, (query, target) in enumerate(parse_alignments(align_file)):
graph.add_edge(query, target)
if c % 100 == 0:
print >> sys.stderr, '...', c
# nx.draw(graph)
# plt.show()
# print graph.nodes()
logfile = open('assemgraph.log', 'w')
visited_nodes = set()
cluster_no = 0
for node in graph.nodes():
if node not in visited_nodes:
filename1 = 'cluster_%d_targets' % cluster_no
filename2 = 'cluster_%d_queries' % cluster_no
ofile1 = open(filename1, 'w')
ofile2 = open(filename2, 'w')
print >> sys.stderr, \
'Writing cluster %d to a file...' % cluster_no
vnodes, max_length = (write_sequence(node, graph, targets, queries,
ofile1, ofile2))
visited_nodes.update(vnodes)
for n in vnodes:
size = len(targets[n]) if n in targets else len(queries[n])
print >> logfile, 'cluster_%d\t%s\t%d' % (cluster_no, n, size)
ofile1.close()
ofile2.close()
print >> sys.stderr, 'total nodes = %d' % len(vnodes)
cluster_no += 1
print >> logfile, '***finished***'
logfile.close()
def main():
try:
input_file = sys.argv[1]
fasta_file = sys.argv[2]
except IndexError:
print >> sys.stderr, \
'unmapped_seq2.py <text file> <fasta file> [min length=50]'
try:
min_length = int(sys.argv[3])
except IndexError:
min_length = 50
db = seqdb.SequenceFileDB(fasta_file)
input_sequences = set()
print >> sys.stderr, 'Reading sequences...',
for line in open(input_file):
input_sequences.add(line.strip())
print >> sys.stderr, 'total %d' % len(input_sequences)
print >> sys.stderr, 'Writing unmapped sequences...'
for seq in db:
sequence = db[seq]
if (seq not in input_sequences and len(sequence) >= min_length):
print >> sys.stderr, 'Writing %s, %d bp' % (seq, len(sequence))
sequtil.write_fasta(sys.stdout, str(sequence), id=seq)
def pygr_slice(fname):
fasta = seqdb.SequenceFileDB(fname)
for rec in fasta:
seq = fasta[rec]
for i in range(NSLICE):
sub = str(seq[:100])
break
def parse_motif():
peak_dict = count_peak()
genome = seqdb.SequenceFileDB('/u/home/f/frankwoe/nobackup/hg19/hg19.noRand.fa')
# motifs include: GGACT, RRACT, DRACH
motif_list = ["(GGACT)", "([GA][GA]ACT)", "([AGT][AG]AC[ACT])"]
with gzip.GzipFile('peak_sum.tsv.gz', 'wb') as f:
# header line
f.write("\t".join( [ 'peak', 'exp_num', 'exp', 'occurence_num', 'occurence', 'fc', 'GGACT,RRACT,DRACH', 'seq' ] ) + '\n')
for peak in peak_dict:
chrom, start, end, strand = peak.split(':')
peak_seq = _fetch_seq(genome, chrom, start, end, strand)
occurence = peak_dict[peak].keys()
occured_exp = list(set([x.split('.')[1] for x in occurence ]))
fc = [ str(peak_dict[peak][x]) for x in occurence ]
motif_count = _match_motif(peak_seq, motif_list)
f.write('\t'.join( [
peak,
str(len(occured_exp)),
','.join(occured_exp),
str(len(occurence)),
','.join(occurence),
','.join(fc),
','.join(motif_count),
peak_seq ]) + '\n' )
def pygr_reverse_comp(fname):
fasta = seqdb.SequenceFileDB(fname)
keys = fasta.keys()
keys.sort()
for rec in keys:
# force full reverse complement
seq = str(-fasta[rec])
sub = seq[:10]
def setUp(self):
genomeFile = '/Users/Likit/projects/mdv/data/chick.fa'
self.genome = seqdb.SequenceFileDB(genomeFile, verbose=False)
exons = [('chr1', 51035309, 51035430), ('chr1', 51062489, 51062516)]
self.isoform = genebuilder.Isoform('chr1', '1', '0', exons, self.genome)
def main():
'''Main function'''
genome = seqdb.SequenceFileDB(sys.argv[1]) # genome sequence
try:
infile = open(sys.argv[2]) # splice sites file from gimme/compare_junction.py
except IndexError, IOError:
# no input file given or cannot open the file,
# read data from stdin instead
infile = sys.stdin
def bed2pygr(dbprefix, referencefile, bedfile, indir):
collision_counter = defaultdict(int)
chrdb = seqdb.SequenceFileDB(referencefile)
annodb = annotation.AnnotationDB({}, chrdb)
al = cnestedlist.NLMSA(dbprefix, 'w', pairwiseMode=True)
load_bed(al, annodb, bedfile, collision_counter)
al.build(saveSeqDict=True)
genomeprefix = os.path.basename(referencefile).rsplit('.', 1)[0]
print >> open(os.path.join(dbprefix) + '.genome', 'w'), genomeprefix
def read_genbank_annots(gbfile, fastafile=None, featureType='CDS',
geneQualifier='gene'):
'''construct annotation DB for gene CDS intervals.
NB: this assumes each gene consists of ONE interval.
This cannot be used for multi-exon genes!'''
try:
gbparse = SeqIO.parse(gbfile, 'genbank')
except TypeError: # SeqIO changed its interface?
ifile = open(gbfile)
try:
gbparse = SeqIO.parse(ifile, 'genbank')
gbseqs = list(gbparse)
finally:
ifile.close()
else:
gbseqs = list(gbparse)
if fastafile is None:
fastafile = gbfile.split('.')[0] + '.fna'
genome = seqdb.SequenceFileDB(fastafile)
genomeIndex = blast.BlastIDIndex(genome) # handle NCBI ID blobs properly
annodb = annotation.AnnotationDB({}, genome,
sliceAttrDict=dict(id=0, start=1, stop=2,
orientation=3))
i = 0
for s in gbseqs:
seqID = genomeIndex[s.id].id # find the right seq and get its actual ID
for f in s.features:
if f.type == featureType:
try:
name = f.qualifiers[geneQualifier][0]
except KeyError: # keep the annotation even if label missing
warnings.warn('Missing gene qualifier "%s" on %s annotation'
% (geneQualifier, featureType))
name = 'unlabeled_%s_%d' % (featureType, i)
i += 1
annodb.new_annotation(name,
(seqID, f.location.start.position,
f.location.end.position, f.strand))
al = cnestedlist.NLMSA('tmp', 'memory', pairwiseMode=True)
for a in annodb.itervalues():
al.addAnnotation(a)
al.build()
return annodb, al, genome
'''
import sys
from pygr import seqdb, sequtil
if len(sys.argv) < 4:
print >> sys.stderr, \
'Usage: split_sequence.py fasta_file chunk_size overlap_size'
raise SystemExit
input_file = sys.argv[1]
chunk_size = int(sys.argv[2])
overlap_size = int(sys.argv[3])
db = seqdb.SequenceFileDB(input_file)
for seq in db:
window = 0
print >> sys.stderr, 'Splitting %s...' % (seq)
if len(db[seq]) <= chunk_size:
sequtil.write_fasta(sys.stdout, str(db[seq]), id=seq)
else:
seq = db[seq]
_id = 1
chunk_id = "%s_%d" % (seq.id, _id)
while window < len(seq):
chunk = seq[window:window + chunk_size]
sequtil.write_fasta(sys.stdout, str(chunk), id=chunk_id)
_id += 1
window += (chunk_size - overlap_size)
def main():
if len(sys.argv) < 4: raise SystemExit
try:
MIN_SCORE = float(sys.argv[5])
except IndexError:
pass
print >> sys.stderr, 'Reading sequence databases...'
queries = seqdb.SequenceFileDB(sys.argv[1])
targets = seqdb.SequenceFileDB(sys.argv[2])
print >> sys.stderr, len(queries), len(targets)
try:
align_file = open(sys.argv[3])
except IOError as e:
print >> sys.stderr, 'Error: check alignment file.'
raise e
aligndb = cnestedlist.NLMSA('alignment', mode='memory', pairwiseMode=True)
print >> sys.stderr, 'Adding sequences to an alignment database...'
# for n, target in enumerate(targets):
# aligndb += targets[target]
# if n % 1000 == 0: print >> sys.stderr, '...', n
target_list = set()
for c, al in enumerate(parse_alignments(align_file)):
aligndb += targets[al.target]
target_list.add(al.target)
add_alignment(aligndb, al, targets, queries)
if c % 100 == 0: print >> sys.stderr, '...', c
print >> sys.stderr, 'Building the alignment database...'
aligndb.build()
print >> sys.stderr, 'Constructing alignment graphs...'
graph = nx.Graph()
for c, target in enumerate(target_list):
try:
sub_ival = targets[target]
for src, dest, edge in aligndb[sub_ival].edges():
source = repr(src).split('[')[0].lstrip('-')
destination = repr(dest).split('[')[0].lstrip('-')
graph.add_edge(source, destination)
except KeyError:
pass
if c % 100 == 0: print >> sys.stderr, '...', c
# nx.draw(graph)
# plt.show()
# print graph.nodes()
logfile = open('assemgraph.log', 'w')
visited_nodes = set()
cluster_no = 0
for node in graph.nodes():
if node not in visited_nodes:
filename1 = 'cluster_%d_targets' % cluster_no
filename2 = 'cluster_%d_queries' % cluster_no
ofile1 = open(filename1, 'w')
ofile2 = open(filename2, 'w')
print >> sys.stderr, \
'Writing cluster %d to a file...' % cluster_no,
vnodes, max_length = (write_sequence(node, graph, targets, queries,
ofile1, ofile2))
visited_nodes.update(vnodes)
for n in vnodes:
size = len(targets[n]) if n in targets else len(queries[n])
print >> logfile, 'cluster_%d\t%s\t%d' % (cluster_no, n, size)
ofile1.close()
ofile2.close()
print >> sys.stderr, '\ttotal nodes = %d' % len(vnodes)
cluster_no += 1
print >> logfile, '***finished***'
logfile.close()
def read_genome(fn):
'''the input of this function should be a FASTA file'''
genome = seqdb.SequenceFileDB(fn)
return genome
'''Get a part of a target sequence that aligned to a given query sequence.'''
import sys
import csv
from pygr import seqdb, sequtil
psl_file = sys.argv[1]
genome_file = sys.argv[2]
genome = seqdb.SequenceFileDB(genome_file)
reader = csv.reader(open(psl_file), dialect='excel-tab')
for cols in reader:
target = cols[13]
start = int(cols[15])
end = int(cols[16])
seq = genome[target][start:end]
seqid = target + '_' + cols[9]
sequtil.write_fasta(sys.stdout, str(seq), id=seqid)
(o, args) = opts.parse_args()
GC_width = int(o.gc_width)
if(o.pygr_seq != None):
if(o.pygr_seq.upper() == "HG18"):
seqs = pygr.Data.Bio.Seq.Genome.Human.hg18()
elif(o.pygr_seq.upper() == "HG19"):
seqs = pygr.Data.Bio.Seq.Genome.Human.hg19()
elif(o.pygr_seq.upper() == "CHIMPY"):
seqs = pygr.Data.Bio.Seq.Genome.chimp.chrY()
elif(o.pygr_seq.upper() == "BACS" or o.pygr_seq.upper() == "CONTROL_BACS"):
seqs = pygr.Data.Bio.Seq.Genome.Human.control_bacs()
elif(o.fn_fastq_seq!=None):
seqs = seqdb.SequenceFileDB(o.fn_fastq_seq)
else:
print("no sequence file input... exiting")
sys.exit(1)
GC_content = {}
grp = "GC_content"
GC_DT = create_gc_DenseTrackSet(o.fnoutTable,o.fnContigLengths,grp,o.overwrite)
for contig in seqs:
if(contig in GC_DT[grp]):
print("loading %s..."%(contig))
#if(chr=="chr21"):
#GC_content["chr21"] = get_chr_correction(hg18,"chr21",GC_width)
def pygr_parse_fasta(fname):
fasta = seqdb.SequenceFileDB(fname)
def main():
infile = sys.argv[1]
genome = seqdb.SequenceFileDB(sys.argv[2])
get_sequence(infile, genome)
def pygr_iter(fname):
fasta = seqdb.SequenceFileDB(fname)
for rec in fasta:
seq = fasta[rec]
def main(options, args):
exons = {}
clusters = {}
newClusterID = 0
clusterConnections = {}
linkedExons = {}
exonPositions = {}
endExons = {}
singleton = 0
print >> sys.stderr, 'Minimum UTR length = ', options.minimumUTRLength
print >> sys.stderr, 'Parsing and clustering exons..'
for n, alnObj in enumerate(psl_parser.read(open(options.infile), 'track')):
tStarts = alnObj.attrib['tStarts']
blockSizes = alnObj.attrib['blockSizes']
if len(blockSizes) == 1:
singleton += 1
tName = alnObj.attrib['tName']
newClusterID = construct(tName, tStarts, blockSizes, exons, clusters,
newClusterID, clusterConnections, linkedExons,
exonPositions, endExons)
if n % 1000 == 0:
print >> sys.stderr, '...', n
print >> sys.stderr, 'Total singletons = ', singleton
sumExons = {}
for ref, end in exons:
try:
sumExons[ref] += 1
except KeyError:
sumExons[ref] = 1
for ref in sorted(sumExons):
print >> sys.stderr, '\t%s has %d exon(s).' % (ref, sumExons[ref])
print >> sys.stderr, '\nTotal %d cluster(s) found.' % len(clusters)
print >> sys.stderr, '\nMerging clusters..'
mergedClusters = mergeClusters(clusters, clusterConnections)
print >> sys.stderr, '\nCleaning up..'
ignored = set([])
for cl in mergedClusters:
allExons = mergedClusters[cl]
cleanUpLinkedExons(allExons, linkedExons, exonPositions, ignored,
options.minimumUTRLength)
print >> sys.stderr, 'Modifying the right end of each transcript..'
for cl in mergedClusters:
findLongestEnd(mergedClusters[cl], linkedExons, endExons,
exonPositions, ignored)
print >> sys.stderr, '\nConstructing transcripts..'
allPaths = {}
visited = set([])
for n, cl in enumerate(mergedClusters):
txExons = sorted(mergedClusters[cl])
paths = buildPaths(linkedExons, txExons, allPaths, ignored, visited)
allPaths[cl] = paths
if n % 1000 == 0:
if n > 0:
print >> sys.stderr, '... %d built..' % n
genome = seqdb.SequenceFileDB(options.genome, verbose=False)
'''Create isoform objects from allPaths and
search for ORF.
'''
print >> sys.stderr, '\nBuilding gene models..'
allGenes = {}
n = 0
for chrom, geneID in allPaths:
n += 1
isoformID = 0
for isoExons in allPaths[(chrom, geneID)]:
isoform = Isoform(chrom, geneID, isoformID, isoExons, genome)
if chrom not in allGenes:
allGenes[chrom] = {}
allGenes[chrom][geneID] = [isoform]
else:
try:
allGenes[chrom][geneID].append(isoform)
except KeyError:
allGenes[chrom][geneID] = [isoform]
isoformID += 1
if n % 1000 == 0:
print >> sys.stderr, '...', n
print >> sys.stderr, '\nRemoving redundant sequences..'
findRedundantSequence(allGenes)
'''Creating sequence records for each DNA, RNA and protein sequences.'''
isoformDNASeqs = []
isoformProteinSeqs = []
isoformRNASeqs = []
totalGenes = 0
for chrom in allGenes:
for geneID in allGenes[chrom]:
totalGenes += 1
isoformID = 0
for isoform in allGenes[chrom][geneID]:
if not isoform.redundant:
isoform.isoformID = isoformID
isoformName = '%s:%d.%d' % (chrom, geneID,
isoform.isoformID)
DNARecord = SeqRecord(isoform.dnaSeq, id=isoformName)
isoformDNASeqs.append(DNARecord)
'''Search for ORF for non-redundant sequences'''
print >> sys.stderr, 'searching ORF: %s:%d.%d' \
% (chrom, geneID,isoformID)
findORF(isoform)
if isoform.frame:
proteinRecord = SeqRecord(isoform.proteinSeq,
id=isoformName)
RNARecord = SeqRecord(isoform.mrnaSeq, id=isoformName)
isoformProteinSeqs.append(proteinRecord)
isoformRNASeqs.append(RNARecord)
isoformID += 1
if n > 0 and n % 1000 == 0:
print >> sys.stderr, '...', n, 'transcripts done.'
print >> sys.stderr, 'Total genes = %d\n\n', totalGenes
print >> sys.stderr, 'Writing gene models to file...'
writeBEDFile(allGenes, options.basename)
print >> sys.stderr, 'Writing DNA sequences to file...'
SeqIO.write(isoformDNASeqs, options.basename + '.dnas.fa', 'fasta')
print >> sys.stderr, 'Writing RNA sequences to file...'
SeqIO.write(isoformRNASeqs, options.basename + '.mrnas.fa', 'fasta')
print >> sys.stderr, 'Writing protein sequences to file...'
SeqIO.write(isoformProteinSeqs, options.basename + '.proteins.fa', 'fasta')
def setUp(self):
hbb1_mouse = testutil.datafile('hbb1_mouse.fa')
self.dna = seqdb.SequenceFileDB(hbb1_mouse)
self.tdb = translationDB.get_translation_db(self.dna)
