def infer_longest_peptide(masses):
'''Returns the longest protein string that matches the spectrum graph of the given masses.'''
# Build the graph from the given masses.
graph = dict()
protein_weight_dict = ProteinWeightDict()
for i in xrange(len(masses)):
for j in xrange(i+1, len(masses)):
# Break the inner loop if we've exceeded the maximum weight.
if masses[j] - masses[i] > max(protein_weight_dict.values()) + 1:
break
# Check if the weight associated with masses i and j approximately matches a known protein.
temp_protein = find_weight_match(masses[j] - masses[i], 0.001)
if temp_protein is not None:
graph[masses[i], masses[j]] = temp_protein
# Get the topological ordering of the graph.
top_order = topological_ordering(graph.keys())
# Build the longest path to each node.
S = {node: '' for node in top_order}
for node in top_order:
for predecessor in map(lambda n: n[0], filter(lambda e: e[1] == node, graph.keys())):
if len(S[predecessor]) + 1 > len(S[node]):
S[node] = S[predecessor] + graph[(predecessor, node)]
# Return the longest path.
return max(S.values(), key=len)
def infer_longest_peptide(masses):
'''Returns the longest protein string that matches the spectrum graph of the given masses.'''
# Build the graph from the given masses.
graph = dict()
protein_weight_dict = ProteinWeightDict()
for i in xrange(len(masses)):
for j in xrange(i + 1, len(masses)):
# Break the inner loop if we've exceeded the maximum weight.
if masses[j] - masses[i] > max(protein_weight_dict.values()) + 1:
break
# Check if the weight associated with masses i and j approximately matches a known protein.
temp_protein = find_weight_match(masses[j] - masses[i], 0.001)
if temp_protein is not None:
graph[masses[i], masses[j]] = temp_protein
# Get the topological ordering of the graph.
top_order = topological_ordering(graph.keys())
# Build the longest path to each node.
S = {node: '' for node in top_order}
for node in top_order:
for predecessor in map(lambda n: n[0],
filter(lambda e: e[1] == node, graph.keys())):
if len(S[predecessor]) + 1 > len(S[node]):
S[node] = S[predecessor] + graph[(predecessor, node)]
# Return the longest path.
return max(S.values(), key=len)
def spectrum(peptide): '''Returns the linear spectrum of a given peptide.''' # Dictionary translating RNA to Protein weight = ProteinWeightDict() # Initialize as the mass 0 and the mass of the entire peptide. spec = [0, sum([int(weight[protein]) for protein in peptide])] # Find the masses of the adjacent intermediary subpeptides spec += [sum([int(weight[protein]) for protein in peptide[j:j+i]]) for i in xrange(1,len(peptide)) for j in xrange(len(peptide)-i+1)] # Sort the list in ascending order and convert to strings. spec = map(str,sorted(spec)) return spec
def cyclospectrum(peptide):
# Dictionary translating RNA to Protein
weight = ProteinWeightDict()
# Initialize as the mass 0 and the mass of the entire peptide.
cyclospec = [0, sum([int(weight[protein]) for protein in peptide])]
# Find the masses of the adjacent intermediary subpeptides
cyclospec += [
sum([int(weight[protein]) for protein in (peptide * 2)[j:j + i]])
for i in xrange(1, len(peptide)) for j in xrange(len(peptide))
]
# Sort the list in ascending order and convert to strings.
cyclospec = map(str, sorted(cyclospec))
return cyclospec
# -*- coding: utf-8 -*-
"""
Created on Tue May 31 19:32:46 2016
@author: Johnqiu
"""
"""
问题1：真实的数据是按大小的顺序排列，模拟数据需不要排序？
"""
from scripts import ProteinWeightDict, IonTypeDict
import random
from operator import itemgetter
aa_table = ProteinWeightDict()
ion_table = IonTypeDict()
def simulatePeptide(pep_len):
acids = [k for k in aa_table]
peptide=[random.choice(acids) for i in range(pep_len)]
peptide = ''.join(peptide) # change list to String
return peptide
def generateSpectrum(peptide, ion_table,intensity = 100):
"""
Args:
-peptide: a peptide string
-iontables: {ion:(offset,prob)}
"""
spectrum = []
prefix_mass = 0
#!/usr/bin/env python
'''
A solution to a ROSALIND bioinformatics problem.
Problem Title: Calculating Protein Mass
Rosalind ID: PRTM
Rosalind #: 020
URL: http://rosalind.info/problems/prtm/
'''
from scripts import ProteinWeightDict
# Load the data.
file1 = open('data/rosalind_prtm.txt')
protein_str = file1.read().strip()
file1.close()
# Load the dictionary that translates protein to monoisotipic weight.
weight_dict = ProteinWeightDict()
# Calculate the weight protein by protein.
monoisotopic_weight = 0
for protein in protein_str:
monoisotopic_weight += weight_dict[protein]
# Print and save the weight.
print monoisotopic_weight
with open('output/020_PRTM.txt', 'w') as output_data:
output_data.write(str(monoisotopic_weight))
def spectrum_score(peptide, exp_spec):
'''Returns the number of matching masses from the spectrum of peptide when compared with the spectrum exp_spec.'''
pep_spec = spectrum(peptide)
# Return -1 if the peptide has more mass than exp_spec.
if pep_spec[-1] > exp_spec[-1]:
return -1
return sum([min(pep_spec.count(protein),exp_spec.count(protein)) for protein in set(pep_spec)])
if __name__ == '__main__':
with open('data/stepic_2e.txt') as input_data:
n, spec = [int(line.strip()) if i==0 else map(int,line.strip().split()) for i, line in enumerate(input_data.readlines())]
# Create the protein weight dictionary.
weight = ProteinWeightDict()
# Initialize the scores dictionary.
scores = dict()
# Build the intial peptides.
seq = filter(lambda L: L[0] != -1, [[spectrum_score(peptide,spec), peptide] for peptide in append_protein(weight.keys())])
# Build the sequence until the masses all grow too large.
while seq != []:
# Store the scores of the current sequence in a dictionary.
scores = dict()
for item in seq:
if item[0] in scores:
scores[item[0]].append(item[1])
else:
scores[item[0]] = [item[1]]
# Dictionary translating RNA to Protein
weight = ProteinWeightDict()
# Initialize as the mass 0 and the mass of the entire peptide.
spec = [0, sum([int(weight[protein]) for protein in peptide])]
# Find the masses of the adjacent intermediary subpeptides
spec += [sum([int(weight[protein]) for protein in peptide[j:j+i]]) for i in xrange(1,len(peptide)) for j in xrange(len(peptide)-i+1)]
# Sort the list in ascending order and convert to strings.
spec = map(str,sorted(spec))
return spec
with open('data/textbook/rosalind_2d.txt') as input_data:
cyclospec = input_data.read().strip().split()
# Create the protein weight dictionary.
weight = ProteinWeightDict()
# Let n be the length of a given peptide, and L be the length of its cyclospectrum. Then L = n(n-1) + 2.
# Using the quadratic formula to to solve for n: n = (sqrt(4L-7) + 1)/2
n = int((sqrt(4*len(cyclospec)-7)+1)/2)
# Find the first n protein in the peptide.
# Need to be careful: two small proteins can add to be less than a larger one, so we can't just take the first n nonzero entries.
# Fortunately, no two small proteins masses add to that of a larger protein.
protein, i = [], 1
while len(protein) != n:
if int(cyclospec[i]) in map(int,weight.values()):
protein.append(cyclospec[i])
i += 1
# Get the name of each protein corresponding to a given weight (if multiple, only take one).
# Find the masses of the adjacent intermediary subpeptides
spec += [
sum([int(weight[protein]) for protein in peptide[j:j + i]])
for i in xrange(1, len(peptide)) for j in xrange(len(peptide) - i + 1)
]
# Sort the list in ascending order and convert to strings.
spec = map(str, sorted(spec))
return spec
with open('data/textbook/rosalind_2d.txt') as input_data:
cyclospec = input_data.read().strip().split()
# Create the protein weight dictionary.
weight = ProteinWeightDict()
# Let n be the length of a given peptide, and L be the length of its cyclospectrum. Then L = n(n-1) + 2.
# Using the quadratic formula to to solve for n: n = (sqrt(4L-7) + 1)/2
n = int((sqrt(4 * len(cyclospec) - 7) + 1) / 2)
# Find the first n protein in the peptide.
# Need to be careful: two small proteins can add to be less than a larger one, so we can't just take the first n nonzero entries.
# Fortunately, no two small proteins masses add to that of a larger protein.
protein, i = [], 1
while len(protein) != n:
if int(cyclospec[i]) in map(int, weight.values()):
protein.append(cyclospec[i])
i += 1
# Get the name of each protein corresponding to a given weight (if multiple, only take one).
def append_protein(add_list):
'''Returns a list containing all peptides from add_list with every possible protein suffix.'''
newlist = []
for item in add_list:
newlist += [item + ch for ch in ProteinWeightDict().keys()]
return newlist
min(pep_spec.count(protein), exp_spec.count(protein))
for protein in set(pep_spec)
])
if __name__ == '__main__':
with open('data/stepic_2e.txt') as input_data:
n, spec = [
int(line.strip()) if i == 0 else map(int,
line.strip().split())
for i, line in enumerate(input_data.readlines())
]
# Create the protein weight dictionary.
weight = ProteinWeightDict()
# Initialize the scores dictionary.
scores = dict()
# Build the intial peptides.
seq = filter(lambda L: L[0] != -1,
[[spectrum_score(peptide, spec), peptide]
for peptide in append_protein(weight.keys())])
# Build the sequence until the masses all grow too large.
while seq != []:
# Store the scores of the current sequence in a dictionary.
scores = dict()
for item in seq:
if item[0] in scores:
scores[item[0]].append(item[1])
else:
def find_weight_match(approx_weight, error):
for item in ProteinWeightDict().items():
if abs(item[1] - approx_weight) < error:
return item[0]
return None
Problem Title: Inferring Protein from Spectrum
Rosalind ID: SPEC
Rosalind #: 053
URL: http://rosalind.info/problems/spec/
'''
from scripts import ProteinWeightDict
# The only major issue is that the given values aren't as precise as those in the table.
# Need to find the closest match (or rewrite the weight dictionary with less precision).
with open('data/rosalind_spec.txt') as input_data:
masses = [float(line.strip()) for line in input_data.readlines()]
# Load a list of (protein, weight) pairs.
weight_list = ProteinWeightDict().items()
# Gives the difference between a given weight and the protein at position i in the weight list.
weight_diff = lambda (i, weight): abs(weight - weight_list[i][1])
# Returns the protein whose mass is closest to specified weight.
closest_prot = lambda weight: weight_list[min(zip(range(len(weight_list)), [weight]*len(weight_list)), key=weight_diff)[0]][0]
# Determine each protein.
prot = [closest_prot(masses[i+1]-masses[i]) for i in range(len(masses)-1)]
# Concatonate to get the desired protein.
print ''.join(prot)
with open('output/053_SPEC.txt', 'w') as output_data:
output_data.write(''.join(prot))
def spectrum_score(peptide, exp_spec):
'''Returns the number of matching masses from the spectrum of peptide when compared with the spectrum exp_spec.'''
pep_spec = spectrum(peptide)
# Return -1 if the peptide has more mass than exp_spec.
if pep_spec[-1] > exp_spec[-1]:
return -1
return sum([min(pep_spec.count(protein),exp_spec.count(protein)) for protein in set(pep_spec)])
if __name__ == '__main__':
with open('data/textbook/rosalind_2e.txt') as input_data:
n, spec = [int(line.strip()) if i==0 else map(int,line.strip().split()) for i, line in enumerate(input_data.readlines())]
# Create the protein weight dictionary.
weight = ProteinWeightDict()
# Initialize the scores dictionary.
scores = dict()
# Build the intial peptides.
seq = filter(lambda L: L[0] != -1, [[spectrum_score(peptide,spec), peptide] for peptide in append_protein(weight.keys())])
# Build the sequence until the masses all grow too large.
while seq != []:
# Store the scores of the current sequence in a dictionary.
scores = dict()
for item in seq:
if item[0] in scores:
scores[item[0]].append(item[1])
else:
scores[item[0]] = [item[1]]
