def calculate_final_probability(ancestor, descendant, final_probs, prob_dict,
end_dict):
three_number_dict = {}
three_number_summaries = [
line.strip().split() for line in open('quick_out')
]
for line in three_number_summaries:
if line[1] not in three_number_dict:
three_number_dict[line[1]] = {}
three_number_dict[line[1]][line[0]] = line[2:]
eightmer_list = custom.count_in_base('AAAAAAAA', 4, 'ACGTz')
for eightmer in eightmer_list:
if eightmer not in final_probs[ancestor][descendant]:
final_probs[ancestor][descendant][eightmer] = [[], [0.0]]
for eightmer in eightmer_list:
one_offer_list = list_one_offers(eightmer)
for one_offer in one_offer_list:
change_prob = (eightmer_prob(eightmer, one_offer, prob_dict,
end_dict)) / 2
final_probs[ancestor][descendant][eightmer][1][0] += change_prob
final_probs[ancestor][descendant][one_offer][0].append(
[change_prob,
int(three_number_dict[ancestor][eightmer][0])])
final_probs[ancestor][descendant][eightmer][1].append(
int(three_number_dict[ancestor][eightmer][0]))
for eightmer in eightmer_list:
expected_gains, gain_substrates = 0, 0
for gain_pair in final_probs[ancestor][descendant][eightmer][0]:
expected_gains += gain_pair[0] * gain_pair[1]
gain_substrates += gain_pair[1]
avg_gain_prob = expected_gains / gain_substrates
gain_pair = map(str, [avg_gain_prob, gain_substrates])
loss_pair = final_probs[ancestor][descendant][eightmer][1]
expected_gains = str(expected_gains)
expected_losses = str(loss_pair[0] * loss_pair[1])
loss_pair = map(str, loss_pair)
output_file.write(
eightmer + '\t' +
'\t'.join(three_number_dict[descendant][eightmer][1:]))
output_file.write('\t' + expected_gains + '\t' + expected_losses +
'\t')
output_file.write('\t'.join(gain_pair) + '\t' + '\t'.join(loss_pair) +
'\n')
def calculate_final_probability(ancestor, descendant, final_probs, prob_dict,
end_dict):
three_number_dict = {}
three_number_summaries = [
line.strip().split() for line in open('quick_out')
]
for line in three_number_summaries:
if line[1] not in three_number_dict:
three_number_dict[line[1]] = {}
three_number_dict[line[1]][line[0]] = line[2:]
eightmer_list = custom.count_in_base('AAAAAAAA', 4, 'ACGTz')
for eightmer in eightmer_list:
if eightmer not in final_probs[ancestor][descendant]:
final_probs[ancestor][descendant][eightmer] = [[], [0.0]]
for eightmer in eightmer_list:
one_offer_list = list_one_offers(eightmer)
for one_offer in one_offer_list:
change_prob = eightmer_prob(eightmer, one_offer, prob_dict,
end_dict)
final_probs[ancestor][descendant][eightmer][1][0] += change_prob
final_probs[ancestor][descendant][one_offer][0].append(
[change_prob,
int(three_number_dict[ancestor][eightmer][0])])
final_probs[ancestor][descendant][eightmer][1].append(
int(three_number_dict[ancestor][eightmer][0]))
for eightmer in eightmer_list:
expected_gains, gain_substrates = 0, 0
for gain_pair in final_probs[ancestor][descendant][eightmer][0]:
expected_gains += gain_pair[0] * gain_pair[1]
gain_substrates += gain_pair[1]
avg_gain_prob = expected_gains / gain_substrates
gain_pair = [avg_gain_prob, gain_substrates]
loss_pair = final_probs[ancestor][descendant][eightmer][1]
expected_losses = loss_pair[0] * loss_pair[1]
print eightmer
print 'actual gains, losses:', three_number_dict[descendant][eightmer][
1:]
print 'expected gains, losses:', expected_gains, expected_losses
print 'gain prob and substrates, loss prob and substrates:', gain_pair, loss_pair
return final_probs
def end_probs(prob_dict):
dinucleotides = custom.count_in_base('AA', 4, 'ACGTz')
choices = 'ACGT'
empty_dict = {}
for anc in choices:
if anc not in empty_dict:
empty_dict[anc] = {}
for desc in choices:
if desc not in empty_dict[anc]:
empty_dict[anc][desc] = {'left': [], 'right': []}
end_probs = copy.deepcopy(empty_dict)
for ancestor in dinucleotides:
end_dict = copy.deepcopy(empty_dict)
for descendant in dinucleotides:
if ancestor[0] == descendant[0] or ancestor[1] == descendant[1]:
end_dict[ancestor[1]][descendant[1]]['left'].append(
prob_dict[ancestor][descendant])
end_dict[ancestor[0]][descendant[0]]['right'].append(
prob_dict[ancestor][descendant])
for ancestor_letter in end_dict:
for descendant_letter in end_dict[ancestor_letter]:
for side in end_dict[ancestor_letter][descendant_letter]:
if len(end_dict[ancestor_letter][descendant_letter]
[side]) >= 1:
end_probs[ancestor_letter][descendant_letter][
side].append(
sum(end_dict[ancestor_letter]
[descendant_letter][side]))
for ancestor_letter in end_probs:
for descendant_letter in end_probs[ancestor_letter]:
for side in end_probs[ancestor_letter][descendant_letter]:
current_list = end_probs[ancestor_letter][descendant_letter][
side]
end_probs[ancestor_letter][descendant_letter][side] = sum(
current_list) / len(current_list)
return end_probs
import custom
eightmer_list=custom.count_in_base('AAAAAAAA', 4, 'ACGTz')
status_list=[line.strip().split('\t') for line in open('enrich_depletion_results_lumped_vert_seeds')]
status_list=status_list[1:]
real_mirs={}
def twomer_counts(eightmer):
eightmer_dict={}
for letter_number in range(7):
twomer=eightmer[letter_number:letter_number+2]
if twomer not in eightmer_dict:
eightmer_dict[twomer]=0
eightmer_dict[twomer]+=1
return eightmer_dict
for line in status_list:
joined_line=''.join(line)
if len(line)>4 and ('hsa' in joined_line or 'ptr' in joined_line or 'ggo' in joined_line or 'ppy' in joined_line):
real_mirs[line[1][1:-1]]=[]
print len(real_mirs)
for real_mir in real_mirs:
real_dict=twomer_counts(real_mir)
for eightmer in eightmer_list:
eightmer_dict={}
test_dict=twomer_counts(eightmer)
if test_dict==real_dict and eightmer[1:-1] not in real_mirs:
real_mirs[real_mir].append(eightmer)
print real_mir, len(real_mirs[real_mir])
'3': '2',
'gorilla': '2',
'2': '1',
'orangutan': '1',
'gibbon': '1'
}
descendant_dict = {
'1': ['gibbon', 'orangutan', '2'],
'2': ['3', 'gorilla'],
'3': ['chimp', 'human']
}
gene_dict = cPickle.load(open('final_utr_dictionary_nogaps'))
species_list = ['3', '2', '1']
good_list = ['A', 'C', 'G', 'T', 'a', 'c', 'g', 't']
probability_dict = {}
dinucleotides = custom.count_in_base('AA', 4, 'ACGTz')
prob_dict, final_probs = {}, {}
for ancestor_di in dinucleotides:
prob_dict[ancestor_di] = {}
for descendant_di in dinucleotides:
prob_dict[ancestor_di][descendant_di] = 0.0
for ancestor in species_list:
if ancestor not in final_probs:
final_probs[ancestor] = {}
descendants = descendant_dict[ancestor]
for descendant in descendants:
if descendant not in final_probs[ancestor]:
final_probs[ancestor][descendant] = {}
mutation_dict = {}
for gene in gene_dict:
for letter_number, junk in enumerate(
import custom
import copy
import cPickle
import time
change_coords = ((0, 1), (1, 2), (2, 3), (3, 4), (4, 5), (0, 2), (0, 3))
patterns = set(custom.count_in_base('00000', 2, '01z'))
patterns.discard('00000')
patterns.discard('11111')
def populate_step(current_step, previous_step_dict, previous_ambig_set,
patterns):
current_step_dict, ambig_set = {}, set([])
patterns_copy = copy.deepcopy(patterns)
print len(patterns), current_step
for pattern_number, pattern in enumerate(patterns):
if pattern_number % 200000 == 0:
print pattern_number / float(len(patterns))
for change in change_coords:
types = set(pattern[change[0]:change[1]])
if len(types) == 1:
if '1' in types:
addition = (change[1] - change[0]) * '0'
change_new = change + ('g', 'g')
elif '0' in types:
addition = (change[1] - change[0]) * '1'
change_new = change + ('l', 'l')
new_pattern = pattern[:change[0]] + addition + pattern[
change[1]:]
if current_step == 1: