Skip to content

GitLab

Commit 716433c2 authored by jkvis's avatar jkvis
Browse files
Options

Testing

parent 008ba894
Hide whitespace changes
Inline Side-by-side
Showing with 45 additions and 42 deletions
repeat-extractor.py
View file @ 716433c2
...@@ -94,47 +94,50 @@ for line in lines: ...@@ -94,47 +94,50 @@ for line in lines:
else: else:
sequences[label] = [string.strip()] sequences[label] = [string.strip()]
select = 'D13S317' #select = 'D13S317'
unit_list = ['TATC'] #unit_list = ['TATC']
reference = sequences[select][0] #reference = sequences[select][0]
sample = sequences[select][0] #sample = sequences[select][0]
description, _, _ = describe_repeats(reference, sample, unit_list) #description, _, _ = describe_repeats(reference, sample, unit_list)
print 'l.{}'.format(description) #print 'l.{}'.format(description)
#for sequence in sequences: for sequence in sequences:
# best = 0 best = 0
# for string in sequences[sequence]: for string in sequences[sequence]:
# repeats = short_sequence_repeat_extractor(string, min_length) repeats = short_sequence_repeat_extractor(string, min_length)
# score = 0 score = 0
# for repeat in repeats: for repeat in repeats:
# if repeat.count + 1 >= min_count: if repeat.count + 1 >= min_count:
# score += (repeat.end - repeat.start) * (repeat.end - repeat.start) * (repeat.count + 1) score += (repeat.end - repeat.start) * (repeat.end - repeat.start) * (repeat.count + 1)
# if score > best: if score > best:
# reference = string reference = string
# best = score best = score
# repeats = short_sequence_repeat_extractor(reference, min_length) repeats = short_sequence_repeat_extractor(reference, min_length)
# units = {} units = {}
# for repeat in repeats: for repeat in repeats:
# if repeat.count + 1 >= min_count: if repeat.count + 1 >= min_count:
# units[reference[repeat.start:repeat.end]] = repeat.count + 1 units[reference[repeat.start:repeat.end]] = repeat.count + 1
# unit_list = [] unit_list = []
# for unit in units: for unit in units:
# unit_list.append(unit) unit_list.append(unit)
# print sequence, reference = sequences[sequence][0]
# if best > 0: print sequence + ':',
# print unit_list print reference
# else: if best > 0:
# print 'no repeat unit identified' print 'repeat units:', unit_list
# reference = sequences[sequence][0] else:
# for string in sequences[sequence]: print 'repeat units: []'
# if best > 0: for string in sequences[sequence]:
# description, _, _ = describe_repeats(reference, string, unit_list) rep_start = 1
# else: rep_end = len(reference)
# description = describe_dna(reference, string) if best > 0:
# print 'l.{}'.format(description) description, rep_start, rep_end = describe_repeats(reference, string, unit_list)
# print else:
description = describe_dna(reference, string)
print '{}({}_{}):l.{}'.format(sequence, rep_start, rep_end, description)
print
Markdown is supported
0% or .
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment