def generate_null_distribution(seq_x, seq_y, scoring_matrix, num_trials):
"""
Inputs:
seq_x, seq_y: character strings that share a common alphabet with
scoring_matrix.
scoring_matrix: output of build_scoring_matrix. Dictionary of
dictionaries whose [seq_x[i]][seq_y[j]] value is the score of the
alignment of seq_x[i], seq_y[i].
num_trials: integer number of simulations to run
Output:
scoring_distribution: a list of scores from the simulations.
Randomly shuffle seq_y num_trial times, score the local alignment with
seq_x.
"""
# initialize
scores = []
# run trials
for trial in range(num_trials):
# shuffle seq_y
_seq_y = list(seq_y)
random.shuffle(_seq_y)
rand_y = ''.join(_seq_y)
# compute local alignment of seq_x and random permutation of seq_y
alignment = seq.compute_alignment_matrix(seq_x, rand_y, scoring_matrix, False)
score = seq.compute_local_alignment(seq_x, rand_y, scoring_matrix, alignment)[0]
# update frequency distribution
scores.append(score)
return scores
def generate_null_distribution(seq_x, seq_y, scoring_matrix, num_trials):
"""
Function for question 4
"""
## make a copy of seq_y:
#new_seq_y = ''
#for each_char in seq_y:
# new_seq_y += each_char
list_seq_y = list(seq_y)
scoring_distribution = {}
for dummy_idx in range(num_trials):
#random.shuffle(new_seq_y)
random.shuffle(list_seq_y)
new_seq_y = ''.join(list_seq_y)
align_matrix = project4.compute_alignment_matrix(
seq_x, new_seq_y, scores, False)
local_result = project4.compute_local_alignment(
seq_x, new_seq_y, scores, align_matrix)
if (local_result[0] in scoring_distribution):
scoring_distribution[local_result[0]] += 1
else:
scoring_distribution[local_result[0]] = 1
print dummy_idx
return scoring_distribution
def generate_null_distribution2(seq_x, seq_y, scoring_matrix, num_trials):
# This function does work. I don't understand why balta2ar write it this way by using distr.json
distr = {
} # store the whole distribution {score1: count1, score2: count2, ..., scoren: countn}
raw = [
] # store all the scores: [score1, score2, ..., scoren], could be duplicate
try:
with open('distr.json') as f:
pair = loads(f.read())
return pair['distr'], pair['raw']
except Exception as e:
print('can\'t open file', str(e))
for _ in range(num_trials):
temp = list(seq_y)
shuffle(temp)
rand_y = ''.join(temp)
align_matrix = compute_alignment_matrix(seq_x, rand_y, scoring_matrix,
False)
score, _, _ = compute_local_alignment(seq_x, rand_y, scoring_matrix,
align_matrix)
if score not in distr:
distr[score] = 0
distr[score] += 1
raw.append(score)
with open('distr.json', 'w') as f:
f.write(dumps({'distr': distr, 'raw': raw}))
return distr, raw
def generate_null_distribution(seq_x, seq_y, scoring_matrix, num_trials):
"""
Function for question 4
"""
## make a copy of seq_y:
#new_seq_y = ''
#for each_char in seq_y:
# new_seq_y += each_char
list_seq_y = list(seq_y)
scoring_distribution = {}
for dummy_idx in range(num_trials):
#random.shuffle(new_seq_y)
random.shuffle(list_seq_y)
new_seq_y = ''.join(list_seq_y)
align_matrix = project4.compute_alignment_matrix(seq_x, new_seq_y, scores, False)
local_result = project4.compute_local_alignment(seq_x, new_seq_y, scores, align_matrix)
if (local_result[0] in scoring_distribution):
scoring_distribution[local_result[0]] += 1
else:
scoring_distribution[local_result[0]] = 1
print dummy_idx
return scoring_distribution
def question_1():
'''
First, load the files HumanEyelessProtein and FruitflyEyelessProtein using
the provided code. These files contain the amino acid sequences that form
the eyeless proteins in the human and fruit fly genomes, respectively. Then
load the scoring matrix PAM50 for sequences of amino acids. This scoring
matrix is defined over the alphabet {A,R,N,D,C,Q,E,G,H,I,L,K,M,F,P,S,T,W,Y,
V,B,Z,X,-} which represents all possible amino acids and gaps (the "dashes"
in the alignment).
Next, compute the local alignments of the sequences of HumanEyelessProtein
and FruitflyEyelessProtein using the PAM50 scoring matrix and enter the
score and local alignments for these two sequences below. Be sure to
clearly distinguish which alignment is which and include any dashes ('-')
that might appear in the local alignment.
'''
human_protein = provided.read_protein(provided.HUMAN_EYELESS_URL)
fruitfly_protein = provided.read_protein(provided.FRUITFLY_EYELESS_URL)
scoring_matrix = provided.read_scoring_matrix(provided.PAM50_URL)
alignment_matrix = project4.compute_alignment_matrix(
human_protein, fruitfly_protein, scoring_matrix, False)
local_alignment = project4.compute_local_alignment(human_protein,
fruitfly_protein,
scoring_matrix,
alignment_matrix)
return local_alignment
def generate_null_distribution(seq_x, seq_y, scoring_matrix, num_trials):
'''
Helper function for Question 4
Takes as input two sequences seq_x and seq_y, a scoring matrix
scoring_matrix, and a number of trials num_trials. This function should
return a dictionary scoring_distribution that represents an un-normalized
distribution generated by performing the following process num_trials times:
Generate a random permutation rand_y of the sequence seq_y using
random.shuffle().
Compute the maximum value score for the local alignment of seq_x and rand_y
using the score matrix scoring_matrix.
Increment the entry score in the dictionary scoring_distribution by one.
'''
scoring_distribution = {}
trial = 0
while trial < num_trials:
seq_y_list = list(seq_y)
random.shuffle(seq_y_list)
rand_y = ''.join(seq_y_list)
alignment_matrix = project4.compute_alignment_matrix(
seq_x, rand_y, scoring_matrix, False)
score = project4.compute_local_alignment(seq_x, rand_y, scoring_matrix,
alignment_matrix)
if score[0] not in scoring_distribution:
scoring_distribution[score[0]] = 1
else:
scoring_distribution[score[0]] += 1
trial += 1
print trial
return scoring_distribution
def question1():
"""
Code for quetion 1
"""
human = read_protein(HUMAN_EYELESS_URL)
fruitfly = read_protein(FRUITFLY_EYELESS_URL)
score_mat = read_scoring_matrix(PAM50_URL)
align_mat = compute_alignment_matrix(human, fruitfly, score_mat, False)
result = compute_local_alignment(human, fruitfly, score_mat, align_mat)
return result
def question1():
# QUESTION 1
align_matrix = project4.compute_alignment_matrix(fruitfly_protein, human_protein, scores, False)
local_alignment_eyeless = project4.compute_local_alignment(fruitfly_protein, human_protein, scores, align_matrix)
#
#for each in local_alignment_eyeless:
# print each
#print local_alignment_eyeless[0]
local_human = local_alignment_eyeless[2]
local_fruitfly = local_alignment_eyeless[1]
def generate_null_distribution(seq_x, seq_y, scoring_matrix, num_trials):
distribution = {}
bar = progressbar.ProgressBar(max_value=1000)
for progress in range(num_trials):
bar.update(progress)
rand_y = list(seq_y)
random.shuffle(rand_y)
alignment_matrix = project4.compute_alignment_matrix(seq_x, rand_y, scoring_matrix, False)
score = project4.compute_local_alignment(seq_x, rand_y, scoring_matrix, alignment_matrix)[0]
distribution[score] = distribution.get(score,0) + 1
save_dict(distribution)
return distribution
def question1():
# QUESTION 1
align_matrix = project4.compute_alignment_matrix(fruitfly_protein,
human_protein, scores,
False)
local_alignment_eyeless = project4.compute_local_alignment(
fruitfly_protein, human_protein, scores, align_matrix)
#
#for each in local_alignment_eyeless:
# print each
#print local_alignment_eyeless[0]
local_human = local_alignment_eyeless[2]
local_fruitfly = local_alignment_eyeless[1]
def align_eyeless(scoring_matrix):
"""
compute the local alignment and score of the human eyeless AA sequence and
the drosophila eyeless AA sequence, using the PAM 50 scoring matrix
"""
# load eyeless AA strings
human = read_protein(HUMAN_EYELESS_URL)
drosophila = read_protein(FRUITFLY_EYELESS_URL)
# compute local alignment matrix
la_mtrx = seq.compute_alignment_matrix(human, drosophila, scoring_matrix, False)
# compute local alignment
return seq.compute_local_alignment(human, drosophila, scoring_matrix, la_mtrx)
def generate_null_distribution(seq_x, seq_y, scoring_matrix, num_trials): ''' 1) Generate a random permutation 'rand_y' of the sequence seq_y 2) Compute the maximum value 'score' for the local alignment of seq_x and rand_y using the score matrix 'scoring_matrix' Return local alignment score ''' temp = list(seq_y) random.shuffle(temp) seq_y = ''.join(temp) local_alignment_matrix = project4.compute_alignment_matrix(seq_x, seq_y, scoring_matrix, False) local_alignment = project4.compute_local_alignment(seq_x, seq_y, scoring_matrix, local_alignment_matrix) return local_alignment[0]
def question_1():
human = read_protein(HUMAN_EYELESS_URL)
fly = read_protein(FRUITFLY_EYELESS_URL)
scoring_matrix = read_scoring_matrix(PAM50_URL)
alignment_matrix = project4.compute_alignment_matrix(human, fly, scoring_matrix, False)
answer = project4.compute_local_alignment(human, fly, scoring_matrix, alignment_matrix)
print "score =", answer[0]
print "align human = ", answer[1]
print "align fly = ", answer[2]
return answer[0]
def question_2():
human = read_protein(HUMAN_EYELESS_URL)
fly = read_protein(FRUITFLY_EYELESS_URL)
consensus = read_protein(CONSENSUS_PAX_URL)
scoring_matrix = read_scoring_matrix(PAM50_URL)
alignment_matrix_local = project4.compute_alignment_matrix(human, fly, scoring_matrix, False)
local_aligns = project4.compute_local_alignment(human, fly, scoring_matrix, alignment_matrix_local)
human_local_align = local_aligns[1]
fly_local_align = local_aligns[2]
human_no_dashes = human_local_align.replace('-','')
fly_no_dashes = fly_local_align.replace('-','')
global_matrix_human_consensus = project4.compute_alignment_matrix(human_no_dashes, consensus, scoring_matrix,True)
global_matrix_fly_consensus = project4.compute_alignment_matrix(fly_no_dashes,consensus, scoring_matrix, True)
global_align_human_consensus = project4.compute_global_alignment(human_no_dashes,consensus,scoring_matrix,global_matrix_human_consensus)
align_global_human = global_align_human_consensus[1]
global_align_fly_consensus = project4.compute_global_alignment(fly_no_dashes, consensus,scoring_matrix,global_matrix_fly_consensus)
align_global_fly = global_align_fly_consensus[1]
count_human = 0
count_fly = 0
#print align_global_human
#print align_global_fly
#print consensus
for pair in zip(align_global_human, consensus):
if pair[0] == pair[1]:
count_human += 1.
for pair in zip(align_global_fly,consensus):
if pair[0] == pair[1]:
count_fly += 1.
human_percentage = (count_human / len(align_global_human)) * 100
fly_percentage = (count_fly / len(align_global_fly)) * 100
print "human percentage: ", human_percentage
print "fly percentage: ", fly_percentage
def generate_null_distribution(seq_x, seq_y, scoring_matrix, num_trials):
"""
Calculate a dictionary scoring_distribution that represents
an un-normalized distribution generated by performing the following
process num_trials times:
Generate a random permutation rand_y of the sequence seq_y
using random.shuffle().
Compute the maximum value score for the local alignment of
seq_x and rand_y using the score matrix scoring_matrix.
Increment the entry score in the dictionary scoring_distribution by one.
Parameters
----------
seq_x: str
a sequence
seq_y: str
another sequence
scoring_matrix: dict of dicts
the scoring matrix
num_trials: int
the number of trials
Returns
-------
scoring_distribution: dict
a dictionary scoring_distribution that represents
an un-normalized distribution
"""
scoring_distribution = defaultdict(int)
for _ in range(num_trials):
rand_y = list(seq_y)
shuffle(rand_y)
align_mat = compute_alignment_matrix(seq_x, rand_y,
scoring_matrix, False)
alignment = compute_local_alignment(seq_x, rand_y,
scoring_matrix, align_mat)
score = alignment[0]
scoring_distribution[score] += 1
return scoring_distribution
def question1():
"""
Compute the local alignments of the sequences of HumanEyelessProtein and
FruitflyEyelessProtein using the PAM50 scoring matrix.
"""
# Compute local alignments.
alignment_matrix = project4.compute_alignment_matrix(HUMAN,
FLY,
SCORING_MATRIX,
global_flag=False)
local_alignment = project4.compute_local_alignment(HUMAN, FLY,
SCORING_MATRIX,
alignment_matrix)
align_human = local_alignment[1]
align_fly = local_alignment[2]
print "Human local alignment:", align_human
print "Fruit fly local alignment:", align_fly
print "score:", local_alignment[0]
return (local_alignment[0], align_human, align_fly)
def generate_null_distribution(seq_x, seq_y, scoring_matrix, num_trials):
"""
Returns a dictionary scoring_distribution that represents an
un-normalized distribution based on the given number of trials num_trials.
"""
scoring_distribution = {}
for dummy in range(num_trials):
y_list = list(seq_y)
random.shuffle(y_list)
rand_y = ''.join(y_list)
alignment_matrix = project4.compute_alignment_matrix(
seq_x, rand_y, scoring_matrix, False)
score = project4.compute_local_alignment(seq_x, rand_y, scoring_matrix,
alignment_matrix)[0]
if score in scoring_distribution.keys():
scoring_distribution[score] = scoring_distribution[score] + 1
else:
scoring_distribution[score] = 1
return scoring_distribution
def generate_null_distribution(seq_x, seq_y, scoring_matrix, num_trials):
"""
generate null distribution of amino acid at specific position
:param seq_x: seq_x
:param seq_y: seq_y
:param scoring_matrix: scoring matrix
:param num_trials: number of trials
:return: a dictionary of scoring_distribution
"""
scoring_distr= {}
for i in xrange(1, num_trials+1):
# random seq from seq_y
rand_y = ''.join(random.sample(seq_y, len(seq_y)))
alignment_matrix = student.compute_alignment_matrix(seq_x, rand_y, scoring_matrix, False)
result = student.compute_local_alignment(seq_x, rand_y, scoring_matrix, alignment_matrix)
scoring_distr[i]= result[0]
return scoring_distr
def generate_null_distribution(seq_x, seq_y, scoring_matrix, num_trials):
"""
generate null distribution of amino acid at specific position
:param seq_x: seq_x
:param seq_y: seq_y
:param scoring_matrix: scoring matrix
:param num_trials: number of trials
:return: a dictionary of scoring_distribution
"""
scoring_distr= {}
for i in xrange(1, num_trials+1):
# random seq from seq_y
rand_y = ''.join(random.sample(seq_y, len(seq_y)))
alignment_matrix = student.compute_alignment_matrix(seq_x, rand_y, scoring_matrix, False)
result = student.compute_local_alignment(seq_x, rand_y, scoring_matrix, alignment_matrix)
scoring_distr[i]= result[0]
return scoring_distr
def generate_null_distribution(seq_x, seq_y, scoring_matrix, num_trials):
distr = {
} # store the whole distribution {score1: count1, score2: count2, ..., scoren: countn}
raw = [
] # store all the scores: [score1, score2, ..., scoren], could be duplicate
for _ in range(num_trials):
temp = list(seq_y)
shuffle(temp)
rand_y = ''.join(temp)
align_matrix = compute_alignment_matrix(
seq_x, rand_y, scoring_matrix,
False) # Returns local alignment matrix.
score, _, _ = compute_local_alignment(seq_x, rand_y, scoring_matrix,
align_matrix)
if score not in distr:
distr[score] = 0
distr[score] += 1
raw.append(score)
return distr, raw
def question1And2():
human = read_protein(HUMAN_EYELESS_URL)
fly = read_protein(FRUITFLY_EYELESS_URL)
print(len(human), len(fly))
scoring = read_scoring_matrix(PAM50_URL)
local_align_matrix = compute_alignment_matrix(human, fly, scoring, False)
score, xs, ys = compute_local_alignment(human, fly, scoring,
local_align_matrix)
print('Question 1')
print('The score of the local alignment is: ', score)
print('The sequence for the HumanEyelessProtein is: ', xs)
print('The sequence for the FruitflyEyelessProtein is: ', ys)
print()
print('Question2')
consensus = read_protein(CONSENSUS_PAX_URL)
# Step1: Delete any dashes '-' present in the sequence.
human_nodash = ''.join([x for x in xs if x != '-'])
fly_nodash = ''.join([y for y in ys if y != '-'])
# Step2: Compute the global alignment of this dash-less sequence with the ConsensusPAXDomain sequence.
hc_global_align_matrix = compute_alignment_matrix(human_nodash, consensus,
scoring, True)
fc_global_align_matrix = compute_alignment_matrix(fly_nodash, consensus,
scoring, True)
# Step3: Compare corresponding elements of these two globally-aligned sequences (local vs consensus) and
# compute the percentage of elements in these two sequences that agree
# NOTE: func agreement contains Stpe2 and Step3.
hc_agree = agreement(human_nodash, consensus, scoring,
hc_global_align_matrix)
fc_agree = agreement(fly_nodash, consensus, scoring,
fc_global_align_matrix)
print('Human vs Consensus agree = %s%%' % hc_agree)
print('Fly vs Consensus agree = %s%%' % fc_agree)
def generate_null_distribution(seq_x, seq_y,scoring_matrix, num_trials):
scoring_distribution = {}
scores_list = []
for i in range(num_trials):
temp = list(seq_y)
random.shuffle(temp)
rand_y = ''.join(temp)
align_matrix = project4.compute_alignment_matrix(seq_x, rand_y, scoring_matrix, False)
local_align = project4.compute_local_alignment(seq_x, rand_y, scoring_matrix, align_matrix)
score = local_align[0]
if score not in scoring_distribution:
scoring_distribution[score] = 0
scoring_distribution[score] += 1
scores_list.append(score)
return scoring_distribution, scores_list
def compare(n, nh, nf, alpha, cons, scoring, align):
'''
n: number of trials
nh: number of characters chosen from alpha and assign to x
nf: number of characters chosen from alpha and assing to y
alpha: original string set: alpha = 'ACBEDGFIHKMLNQPSRTWVYXZ'
cons: Consensus strings
scoring: scoring matrix for alpha
align: alignment matrix????? What is this? Somthing wrong??
'''
ag1, ag2 = [], []
for i in range(n):
x, y = rprot(nh, alpha), rprot(nf, alpha)
_, xs, ys = compute_local_alignment(x, y, scoring, align)
xs_nodash = ''.join([x for x in xs if x != '-'])
ys_nodash = ''.join([y for y in ys if y != '-'])
ag1.append(agreement(xs_nodash, cons, scoring, align))
ag2.append(agreement(ys_nodash, cons, scoring, align))
hc_agree = sum(ag1) / float(n)
fc_agree = sum(ag2) / float(n)
print('Random Human vs Consensus agree = %s%%' % hc_agree)
print('Random Fly vs Consensus agree = %s%%' % fc_agree)
A string representing the protein
"""
protein_file = urllib2.urlopen(filename)
protein_seq = protein_file.read()
protein_seq = protein_seq.rstrip()
return protein_seq
# Q1
#compute_local_alignment(seq_x, seq_y, scoring_matrix, alignment_matrix)
seq_fly = read_protein(FRUITFLY_EYELESS_URL)
seq_human = read_protein(HUMAN_EYELESS_URL)
score_matrix = read_scoring_matrix(PAM50_URL)
alignment_matrix = student.compute_alignment_matrix(seq_human, seq_fly, score_matrix, False)
result = student.compute_local_alignment(seq_human, seq_fly, score_matrix, alignment_matrix)
#print result[0]
#human
#print result[1]
#fly
#print result[2]
# Q2
seq_pax = read_protein(CONSENSUS_PAX_URL)
#fly and pax domain
# alignment_matrix_global = student.compute_alignment_matrix(result[2], seq_pax, score_matrix, True)
# result2 = student.compute_global_alignment(result[2], seq_pax, score_matrix, alignment_matrix_global)
#print result2[0]
#print result2[1]
def align_human_fly_protein():
alignment_matrix = project4.compute_alignment_matrix(protein_human, protein_fly, scoring_matrix, False)
result = project4.compute_local_alignment(protein_human, protein_fly, scoring_matrix, alignment_matrix)
return result
# template lines and solution lines list of line string
word_list = words.split('\n')
print "Loaded a dictionary with", len(word_list), "words"
return word_list
# question 1
scoring_matrix = read_scoring_matrix(PAM50_URL)
seq_x = read_protein(HUMAN_EYELESS_URL)
seq_y = read_protein(FRUITFLY_EYELESS_URL)
consensusseq = read_protein(CONSENSUS_PAX_URL)
alignment_matrix = student.compute_alignment_matrix(seq_x, seq_y,
scoring_matrix, False)
score, string_Hu, string_Fr = student.compute_local_alignment(
seq_x, seq_y, scoring_matrix, alignment_matrix)
print string_Hu
newstring_Hu = ""
for elem in string_Hu:
if elem != '-':
newstring_Hu += elem
print newstring_Hu
newstring_Fr = ""
for elem in string_Fr:
if elem != '-':
newstring_Fr += elem
alignment_matrix_Hum_local_Con = student.compute_alignment_matrix(
newstring_Hu, consensusseq, scoring_matrix, True)
score1, str_Hu_Con, str_Con_Hu = student.compute_global_alignment(
def read_words(filename):
"""
Load word list from the file named filename.
Returns a list of strings.
"""
# load assets
word_file = urllib2.urlopen(filename)
# read in files as string
words = word_file.read()
# template lines and solution lines list of line string
word_list = words.split('\n')
print "Loaded a dictionary with", len(word_list), "words"
return word_list
HUMAN_EYELESS_PROTEIN = read_protein(HUMAN_EYELESS_URL)
FRUITFLY_EYELESS_PROTEIN = read_protein(FRUITFLY_EYELESS_URL)
PAM50_SCORING_MATRIX = read_scoring_matrix(PAM50_URL)
PAM50_ALIGNMENT_MATRIX = student.compute_alignment_matrix(
HUMAN_EYELESS_PROTEIN, FRUITFLY_EYELESS_PROTEIN, PAM50_SCORING_MATRIX,
False)
print student.compute_local_alignment(HUMAN_EYELESS_PROTEIN,
FRUITFLY_EYELESS_PROTEIN,
PAM50_SCORING_MATRIX,
PAM50_ALIGNMENT_MATRIX)
"""
protein_file = urllib2.urlopen(filename)
protein_seq = protein_file.read()
protein_seq = protein_seq.rstrip()
return protein_seq
# Q1
#compute_local_alignment(seq_x, seq_y, scoring_matrix, alignment_matrix)
seq_fly = read_protein(FRUITFLY_EYELESS_URL)
seq_human = read_protein(HUMAN_EYELESS_URL)
score_matrix = read_scoring_matrix(PAM50_URL)
alignment_matrix = student.compute_alignment_matrix(seq_human, seq_fly,
score_matrix, False)
result = student.compute_local_alignment(seq_human, seq_fly, score_matrix,
alignment_matrix)
#print result[0]
#human
#print result[1]
#fly
#print result[2]
# Q2
seq_pax = read_protein(CONSENSUS_PAX_URL)
#fly and pax domain
# alignment_matrix_global = student.compute_alignment_matrix(result[2], seq_pax, score_matrix, True)
# result2 = student.compute_global_alignment(result[2], seq_pax, score_matrix, alignment_matrix_global)
#print result2[0]
#print result2[1]
#print result2[2]
