def pattern_matching_mismatches(text, k, d):
counts = defaultdict(int)
max_count = 0
for i in range(len(text) - (k - 1)):
kmer = text[i : i + k]
for n in set.union(neighbours(kmer, d), neighbours(reverse_complement(kmer), d)):
counts[n] += 1
if counts[n] > max_count:
max_count = counts[n]
most_frequent = []
for kmer, count in counts.items():
if count == max_count:
most_frequent.append(kmer)
return most_frequent
def pattern_matching_mismatches_copy_pseudocode(text, k, d):
frequent_patterns = set()
close = []
frequency_array = []
for i in range(4**k):
close.append(False)
frequency_array.append(0)
for i in range(len(text) - (k)):
neighbourhood = neighbours(text[i:i+k], d)
for pattern in neighbourhood:
index = pattern_to_number(pattern)
close[index] = True
for i in range(4**k):
if close[i]:
pattern = number_to_pattern(i, k)
frequency_array[i] = approx_pattern_count(text, pattern, d)
max_count = max(frequency_array)
for i in range(4**k):
if frequency_array[i] == max_count:
pattern = number_to_pattern(i, k)
frequent_patterns.add(pattern)
return frequent_patterns
def motif_enumeration(dna_strings, k, d):
# Epic runtime!
patterns = set()
for dna in dna_strings:
for i in xrange(len(dna) - k + 1):
ns = neighbours(dna[i:i + k], d)
for n in ns:
if hamming_in_all([x for x in dna_strings if x != dna], n, d):
patterns.add(n)
return patterns