def setUp(self):
aligner = PairwiseAligner()
aligner.internal_open_gap_score = -1
aligner.internal_extend_gap_score = -0.0
aligner.match_score = +1
aligner.mismatch_score = -1
aligner.mode = "local"
self.aligner = aligner
def bioPython_default_local_aligner(a, b):
aligner = PairwiseAligner()
aligner.mode = 'local'
aligner.match_score = 2
aligner.mismatch_score = -3
aligner.open_gap_score = -7
aligner.extend_gap_score = -2
sequence1 = SeqIO.read('./resource/fasta' + str(a) + '.fasta', 'fasta')
sequence2 = SeqIO.read('./resource/fasta' + str(b) + '.fasta', 'fasta')
alignments = aligner.align(sequence1.seq, sequence2.seq)
def get_clusters_from_seqlist(seqlist, dist_threshold=0.05):
"""Cluster a list of sequences by a distance identity threshold
Parameters
----------
seqlist : list
list of sequences as str
dist_threshold : float
Max distance value to retain, branches above this length in the
hierarchical clustering tree will be cut.
Returns
-------
list
list of lists - input sequences now grouped by cluster
list
list of int - cluster memberships of the originally input list
"""
if len(seqlist) == 1:
# Skip alignment if there is only one sequence
return([seqlist], [0])
else:
aligner = PairwiseAligner()
aligner.mode = "local"
# Convert sequence list to distance matrix
distmatrix = []
for seq1 in seqlist:
row = []
for seq2 in seqlist:
maxlen = max([len(seq1), len(seq2)])
# Take percentage identity of pairwise alignment score (match base
# +1, all other operations +0) over the longer sequence in pair
idval = aligner.align(seq1, seq2).score / maxlen
distval = 1 - idval # convert to distance fraction
row.append(distval)
distmatrix.append(row)
# Hierarchical clustering from the distance matrix
htree = treecluster(data=None, distancematrix=array(distmatrix))
# Find number of branches with length longer than threshold, and add 1
# to get number of cuts
cuts = 1 + len([htree[i].distance for i in range(len(htree))
if htree[i].distance > dist_threshold])
clust_ids = list(htree.cut(cuts))
clust_seqs_dict = defaultdict(list)
for i in range(len(seqlist)):
clust_seqs_dict[clust_ids[i]] += [seqlist[i]]
# Convert dict of lists to list of lists
clust_seqs = [clust_seqs_dict[i] for i in clust_seqs_dict]
return(clust_seqs, clust_ids)
def pairwise(self, potential_parent):
"""
Парное выравнивание последовательности листа на потенциальных родителей и сохранение скора
Args:
potential_parent (str): последовательность потенциального родителя
"""
aligner = PairwiseAligner()
# используем локальное выравнивание
aligner.mode = 'local'
# заменим дефолтные символы пропуска на другие
seq = self.seq.replace('-', '')
potential_parent_undersores = potential_parent.replace('-', '')
# выравнивание
score = aligner.score(seq, potential_parent_undersores)
# сохраняем скор
self.parent_scores[potential_parent] = score
def perform_randomized_tests(n=1000):
"""Perform randomized tests and compare to pslMap.
Run this function to perform 8 x n mappings for alignments of randomly
generated sequences, get the alignment in PSL format, and compare the
result to that of pslMap.
"""
aligner = PairwiseAligner()
aligner.internal_open_gap_score = -1
aligner.internal_extend_gap_score = -0.0
aligner.match_score = +1
aligner.mismatch_score = -1
aligner.mode = "local"
for i in range(n):
nBlocks1 = random.randint(1, 10)
nBlocks2 = random.randint(1, 10)
test_random(aligner, nBlocks1, nBlocks2, "+", "+")
test_random(aligner, nBlocks1, nBlocks2, "+", "-")
test_random(aligner, nBlocks1, nBlocks2, "-", "+")
test_random(aligner, nBlocks1, nBlocks2, "-", "-")
test_random_sequences("+", "+")
test_random_sequences("+", "-")
test_random_sequences("-", "+")
test_random_sequences("-", "-")
type=str,
required=True)
parser.add_argument('-r',
'--reference',
help='Reference to be aligned to',
type=str,
required=True)
parser.add_argument('-n',
'--seq_name',
help='Name of the aligned sequence',
type=str,
required=True)
args = parser.parse_args()
aligner = PairwiseAligner()
aligner.mode = 'global'
aligner.match_score = 1
aligner.mismatch_score = 0
aligner.open_gap_score = -2
aligner.extend_gap_score = -1
ref = SeqIO.read(args.reference, "fasta")
ref.seq = str(ref.seq.upper()).replace('-', 'N')
cons = SeqIO.read(args.infile, "fasta")
aln = aligner.align(ref.seq, cons.seq)
with open(args.outfile, 'w') as out:
print(">", args.seq_name, file=out)
print(str(aln[0]).strip().split('\n')[2], file=out)
args.open_gap_score = -5
if not args.extend_gap_score:
args.extend_gap_score = -2
else:
sub_matrix = getattr(import_module('Bio.SubsMat.MatrixInfo'),
args.sub_matrix)
aligners['global'].substitution_matrix = sub_matrix
if not args.open_gap_score:
args.open_gap_score = -11
if not args.extend_gap_score:
args.extend_gap_score = -1
aligners['global'].open_gap_score = args.open_gap_score
aligners['global'].extend_gap_score = args.extend_gap_score
if args.sim_algo == 'smith-waterman':
aligners['local'] = PairwiseAligner()
aligners['local'].mode = 'local'
if args.seq_type in ('dna', 'rna'):
aligners['local'].match = args.match_score
aligners['local'].mismatch = args.mismatch_score
else:
aligners['local'].substitution_matrix = sub_matrix
aligners['local'].open_gap_score = args.open_gap_score
aligners['local'].extend_gap_score = args.extend_gap_score
# Karlin-Altschul parameter values
if args.seq_type in ('dna', 'rna'):
if ((args.match_score, args.mismatch_score) in KA_PARAMS['na']
and (abs(args.open_gap_score), abs(args.extend_gap_score))
in KA_PARAMS['na'][(args.match_score, args.mismatch_score)]):
args.ka_gapped_l = KA_PARAMS['na'][(args.match_score,
args.mismatch_score)][(
# # DEMO: Izračun lastne matrike in njena uporaba
#
# V spodnjem primeru je prikazan izračun lastne matrike (enako, kot pri prejšnji vaji - [VAJA: Izračun matrike zamenjav (Python)](matrika_zamenjav.ipynb)) ter uporaba tako izračunane matrike za poravnavo dveh zaporedij.
#
# ## Izračun matrike
# In[1]:
from Bio import SeqIO
sequence1 = SeqIO.read('vhod/matrika_zamenjav-myoglobin_horse.fasta', 'fasta')
sequence2 = SeqIO.read('vhod/matrika_zamenjav-myoglobin_rat.fasta', 'fasta')
# v mapi vhod sta tudi zaporedji mišjega in človeškega nebulina, ki sta bistveno daljši
from Bio.Align import PairwiseAligner
aligner = PairwiseAligner()
aligner.mode = 'local'
aligner.match_score = 2
aligner.mismatch_score = -3
aligner.open_gap_score = -7
aligner.extend_gap_score = -2
alignments = aligner.align(sequence1.seq, sequence2.seq)
alignment = alignments[0]
from Bio.Align.substitution_matrices import Array
frequency = Array('ACGT', dims=2)
for (start1, end1), (start2, end2) in zip(*alignment.aligned):
seq1 = sequence1[start1:end1]
seq2 = sequence2[start2:end2]
for c1, c2 in zip(seq1, seq2):
frequency[c1, c2] += 1
import numpy
probabilities = frequency / numpy.sum(frequency)