def get_seq(self, chrom, start, end, strand):
chrom = misc.match_chrom_format(chrom, list(self.fasta.keys()))
seq = self.fasta[chrom][start:end + 1]
if strand == "-":
seq = misc.reverse_comp(seq)
return seq
def align(self, seq):
alns = []
revseq = reverse_comp(seq)
for i, name in enumerate(self.names_to_aligners):
aligner = self.names_to_aligners[name]
faln = aligner.align(seq)
raln = aligner.align(revseq)
cur_aln = pysam.AlignedSegment()
cur_aln.reference_id = i
if faln.score > raln.score:
cur_aln.query_sequence = seq
cur_aln.reference_start = faln.ref_begin
cur_aln.set_tag("AS", faln.score)
cur_aln.cigarstring = faln.cigar_string
else:
cur_aln.query_sequence = revseq
cur_aln.reference_start = raln.ref_begin
cur_aln.set_tag("AS", raln.score)
cur_aln.cigarstring = raln.cigar_string
cur_aln.is_reverse = True
alns.append(cur_aln)
return alns
def simulate_read_pair(sequence, start, length=150, isize=400, flip=False):
r1 = pysam.AlignedSegment()
r1.query_sequence = sequence[start:start + length]
r2 = pysam.AlignedSegment()
pos2 = start + isize
r2.query_sequence = reverse_comp(sequence[pos2 - length:pos2])
if flip:
r1, r2 = r2, r1
return ReadPair(Alignment(r1), Alignment(r2), read_stats)
def simple_dotplot(s1, s2, wordsize=8, scale=650):
# scale is the final size of the output matrix for visualization
l1 = int((len(s1)-wordsize))
l2 = int((len(s2)-wordsize))
width = int(numpy.ceil(l1/max([l1,l2]) * scale))
height = int(numpy.ceil(l2/max([l1,l2]) * scale))
mat = numpy.zeros((height, width))
binsize = l1/(width-1)
kmertopos1 = collections.defaultdict(list)
# get positions of kmers in s1
for i in range(l1):
kmer = s1[i:i+wordsize]
kmertopos1[kmer].append(i)
# find all matching kmers from s2
for i in range(l2):
kmer = s2[i:i+wordsize]
positions = kmertopos1[kmer]
positions = (numpy.array(positions)/binsize).astype(int)
y = int(i/binsize)
mat[y, positions] += 1
# find all rev-comp kmer matches from s2
for i in range(l2):
kmer = misc.reverse_comp(s2[i:i+wordsize])
positions = kmertopos1[kmer]
positions = (numpy.array(positions)/binsize).astype(int)
y = int(i/binsize)
mat[y, positions] += 1
mat = mat[::-1,]
return mat
def get_seq(self, chrom, start, end, strand):
seq = self.names_to_contigs[chrom][start:end + 1]
if strand == "-":
seq = misc.reverse_comp(seq)
return seq
def original_sequence(self):
if self.is_reverse:
return misc.reverse_comp(self.query_sequence)
return self.query_sequence
