def edit_distance(seq_x, seq_y, scoring_matrix):
"""
calculate the edit distance through seq_x, seq_y and scoring matrix
by return |seq_x| + |seq_y| - score of the corresponding global alignment
"""
alignment_matrix = pj4.compute_alignment_matrix(seq_x, seq_y, scoring_matrix, True)
score, align_x, align_y = pj4.compute_global_alignment(seq_x, seq_y, scoring_matrix, alignment_matrix)
return len(seq_x) + len(seq_y) - score
def ED_xcross_test(diag_score, off_diag_score, dash_score):
"""
Insertion: abc -> abbc
Deletion: abc -> ac
Subsititution: abc -> abd
"""
scoring_matrix = pj4.build_scoring_matrix({"a", "b", "c", "d"}, diag_score, off_diag_score, dash_score)
# test I: x = 'abcd', y = 'ad', the edit distance is 4
test1_ED = edit_distance("ab", "acccc", scoring_matrix)
# test II: x = 'abc', y = 'abcddd', the edit distance is 3
test2_ED = edit_distance("acccd", "ad", scoring_matrix)
# test III: x = 'abcd', y = 'abb', the edit distance is 2
test3_ED = edit_distance("abcd", "addd", scoring_matrix)
return (test1_ED, test2_ED, test3_ED)
def check_spelling(checked_word, dist, word_list):
"""
To iterates through word_list and returns the set of
all words that are within edit distance dist of the string checked_word
"""
diag_score = 2
off_diag_score = 1
dash_score = 0
scoring_matrix = pj4.build_scoring_matrix(set('qazwsxedcrfvtgbyhnujmikolp'), diag_score, off_diag_score, dash_score)
word_list = set(word_list)
candidate_words = list()
count = 0
for word in word_list:
if len(word) < len(checked_word) - dist or len(word) > len(checked_word) + dist:
continue
# number of operation = 2
# 2 insertion
passed = False
for number in range(len(checked_word)):
if checked_word[:number] in word and checked_word[number + 2:] in word:
passed = True
# 1 insertion
passed = True
for number in range(len(checked_word)):
if checked_word[:number] not in word or checked_word[number + 1:] not in word:
passed = False
if not passed:
continue
count += 1
if sol4_7.edit_distance(checked_word, word, scoring_matrix) <= dist:
candidate_words.append(word)
print count
return set(candidate_words)
def generate_null_distribution(seq_x, seq_y, scoring_matrix, num_trials):
"""
To return a dictionary scoring_distribution that represents an
un-normalized distribution generated by performing the following local
alignment process num_trials times.
"""
scoring_distribution = dict()
for dummy_idx in range(num_trials):
tmp_y = list(seq_y)
random.shuffle(tmp_y)
rand_y = ''.join(tmp_y)
alignment_matrix = pj4.compute_alignment_matrix(seq_x, rand_y, scoring_matrix, False)
score = max([max(value) for value in alignment_matrix])
#score, align_x, align_y = pj4.compute_local_alignment(seq_x, rand_y, scoring_matrix, alignment_matrix)
if score not in scoring_distribution.keys():
scoring_distribution[score] = 1
else:
scoring_distribution[score] += 1
return scoring_distribution
""" Algorithm thinking application 4-1 data: 2015/07/30 Author: You-Hao """ import alg_application4_provided as app4 import AT_project_4 as pj4 protein_human = app4.read_protein(app4.HUMAN_EYELESS_URL) protein_fruitfly = app4.read_protein(app4.FRUITFLY_EYELESS_URL) scoring_matrix = app4.read_scoring_matrix(app4.PAM50_URL) alignment_matrix_4_1 = pj4.compute_alignment_matrix(protein_human, protein_fruitfly, scoring_matrix, False) score_4_1, align_human_4_1, align_fruitfly_4_1 = pj4.compute_local_alignment(protein_human, protein_fruitfly, scoring_matrix, alignment_matrix_4_1) print score_4_1 print align_human_4_1 print align_fruitfly_4_1
seq_human_nodash = ''
seq_fruitfly_nodash = ''
for char in seq_human:
if char != '-':
seq_human_nodash = seq_human_nodash + char
for char in seq_fruitfly:
if char != '-':
seq_fruitfly_nodash = seq_fruitfly_nodash + char
print len(seq_human_nodash)
print len(seq_fruitfly_nodash)
# for human
alignment_matrix = pj4.compute_alignment_matrix(seq_human_nodash, seq_PAX, scoring_matrix, True)
score_human, align_human, align_PAX_1 = pj4.compute_global_alignment(seq_human_nodash, seq_PAX, scoring_matrix, alignment_matrix)
print score_human
print align_human
print align_PAX_1
match_human = 0
for ind in range(len(align_human)):
if align_human[ind] == align_PAX_1[ind]:
match_human += 1
print float(match_human) / len(align_human) * 100.
# for fruit fly
alignment_matrix = pj4.compute_alignment_matrix(seq_fruitfly_nodash, seq_PAX, scoring_matrix, True)
