def get_random_tree(filename, tree_string, L, kappa):
# strains = read_in_strains(filename)
# # L = genome_length(strains)
# min_m = get_min_m(strains, L)
# scaled_tree_string = scale_newick_format_tree(strains, L, min_m, tree_string)
phylogeny = pyvolve.read_tree(tree = tree_string)
# pyvolve.print_tree(phylogeny)
freqs = [0.25,0.25,0.25,0.25]
nuc_model = pyvolve.Model('nucleotide', {'kappa':kappa, 'state_freqs':freqs})
ancestor = generate_ancestor(L)
print(ancestor)
my_partition = pyvolve.Partition(models = nuc_model, root_sequence = ancestor)
my_evolver = pyvolve.Evolver(partitions = my_partition, tree = phylogeny)
my_evolver()
# my_evolver(write_anc = True)
simulated_strains = my_evolver.get_sequences()
# strains = my_evolver.get_sequences(anc = True)
# strain_names = list(strains.keys())
pi = pi_value(simulated_strains)
theta = theta_value(simulated_strains)
# print('pi: ' + str(pi))
# print('theta: ' + str(theta))
return {'pi': pi, 'theta': theta}
def get_random_tree(L, species, scaled_tree_string, kappa, iteration):
# strains = read_in_strains(filename)
# L = genome_length(strains)
# min_m = get_min_m(strains, L)
# max_m = get_max_m(strains, L, tree_string)
# pis = []
# thetas = []
# scaled_trees = []
# for x in range(min_m,max_m+1):
# scaled_tree_string = scale_newick_format_tree(strains, L, x, tree_string, increment)
# scaled_trees.append(scaled_tree_string)
# for tree in scaled_trees:
phylogeny = pyvolve.read_tree(tree=scaled_tree_string)
print('read in the tree')
pyvolve.print_tree(phylogeny)
freqs = [0.25, 0.25, 0.25, 0.25]
nuc_model = pyvolve.Model('nucleotide', {
'kappa': kappa,
'state_freqs': freqs
})
ancestor = generate_ancestor(L)
print('generated an ancestor')
# # print(ancestor)
my_partition = pyvolve.Partition(models=nuc_model, root_sequence=ancestor)
my_evolver = pyvolve.Evolver(partitions=my_partition, tree=phylogeny)
my_evolver(ratefile=None,
infofile=None,
seqfile="simulated_alignment_" + str(species[:-1]) +
"_universal_" + str(iteration + 1) + ".fasta")
# # my_evolver()
print('evolved the sequences')
# # my_evolver(write_anc = True)
simulated_strains = my_evolver.get_sequences()
# # strains = my_evolver.get_sequences(anc = True)
# # strain_names = list(strains.keys())
pi = pi_value(simulated_strains)
theta = theta_value(simulated_strains)
# pis.append(pi)
# thetas.append(theta)
# # print('pi: ' + str(pi))
# # print('theta: ' + str(theta))
# return {'pi': pis, 'theta': thetas}
return pi, theta
# with open(('ID_Matrix_' + species_name + '.csv'), 'w', newline = '') as f:
# writer = csv.writer(f)
# writer.writerow([species_name])
# header = ['']
# header.extend(strain_names)
# writer.writerow(header)
# for row in range(shape[0]):
# write_row = [strain_names[row]]
# write_row.extend(identity_matrix[row])
# writer.writerow(write_row)
# writer.writerow([])
print(name)
L = genome_length(strains)
n = species_size(strains)
pi = pi_value(strains)
theta = theta_value(strains)
else:
L = 'N/A'
n = 'N/A'
# if (os.path.join(full_path, 'concat_universal.fa'):
# concat_universal_file = open(os.path.join(full_path, 'concat_universal.fa'), 'r')
# concat_universal_file = list(concat_universal_file)
# print(full_path)
# concat_core_file = open(os.path.join(full_path, 'concat_core.fa'), 'r') # open('concat_core.fa', 'r')
# concat_universal_file = open(os.path.join(full_path, 'concat_universal.fa'), 'r')
full_path = os.path.join(folder_path, 'kappa.txt')
if os.path.exists(full_path):
kappa_file = open(full_path, 'r')
def apply_model_along_phylogeny(species_path, kappa, tree_string):
# ancestor = ''
# print('Reading in the strains.\n\n')
strains = read_in_strains(
species_path
) # dictionary with the genomes of all the strains; key = strain name, value = genome
internal_nodes = get_internal_nodes(anc_path)
all_nodes = strains + internal_nodes
strain_names = list(strains.keys()) # list of all the extant strain names
all_node_names = list(all_nodes.keys())
# print(strain_names)
n = species_size(strains) # number of extant strains
total_pairs = (
n *
(n - 1)) / 2 # the total number of strain pairs that will be compared
L = genome_length(strains) # number of base pairs in the genomd
theta = theta_value(
strains) # proportion of the genome that is polymorphic
mu = theta / (2 * n
) # mutation rate in mutations per base pair per generation
min_m = get_min_m(
strains, L
) # minimum number of mutations that could account for all the polymorphisms in the species
# print('Scaling the branch lengths of the tree.\n\n')
# scaled_tree_string = tree_string
scaled_tree_string = scale_newick_format_tree(
strains, L, min_m, tree_string, 0) # the tree_string scaled by min_m
SHARED = np.empty(
[n, n], dtype=np.float, order='C'
) # a matrix of the number of nucleotides shared between two strains; the (i,j) entry is the number of nucleotides that are the same between strain i and strain j
CONVERGENT = np.empty(
[n, n], dtype=np.float, order='C'
) # a matrix of the number of nucleotides that match due to convergent mutation between two strains; the (i,j) entry is the number of convergent mutations between strain i and strain j
ANCESTRAL = np.empty(
[n, n], dtype=np.float, order='C'
) # a matrix of the number of nucleotides that match due to direct inheritence from the ancestor; the (i,j) entry is the number of nucleotides that were inherited by both strain i and strain j
RECOMBINANT = np.empty(
[n, n], dtype=np.float, order='C'
) # a matrix of the number of nucleotides that match due to a recombination event; the (i,j) entry is the number of nucleotides that were recombined between strain i and strain j
updated_tree_info = name_nodes(tree_string, strain_names)
tree_string = updated_tree_info[
'tree_string'] # version of the tree_string where every node is labeled
new_nodes = updated_tree_info[
'new_nodes'] # the new node names that were added
all_nodes = all_node_names + new_nodes # list of all the node names in the pyhlogenetic tree
# print(all_nodes)
# parents = find_parents(strain_names, tree_string) # a dictionary of the sequence of parents of each strain; key = strain name, value = list of the parents in order of increasing distance from the strain
parents = find_parents(all_node_names, tree_string)
distances = get_branch_lengths(
all_nodes, tree_string
) # a dictionary of the distances of each strain to its closest ancestor; key = strain name, value = distance to its closest ancestor
# print(parents)
# print(distances)
# parents = parents_and_distances['parents']
# distances = parents_and_distances['distances']
count = 1 # a counter for the current strain pair number that is being processed
for s1 in range(n): # allows each strain to be strain 1
strain1 = strain_names[s1]
genome1 = strains[strain1]
SHARED[s1, s1] = L
CONVERGENT[
s1,
s1] = 0 # there can be no convergent mutations between a strain and itself
ANCESTRAL[s1, s1] = L
RECOMBINANT[s1, s1] = 0
for s2 in range(s1 + 1,
n): # allows each strain after strain 1 to be strain 2
strain2 = strain_names[s2]
genome2 = strains[strain2]
MRCA = find_MRCA(
strain1, strain2, parents
) # the Most Recent Common Ancestor between the two strains
s, a = 0, 0 # initializes the shared and ancestral values for the pair of strains to 0
for site in range(L): # goes through every site along the genome
if genome1[site] == genome2[
site]: # counts up the number of shared sites
s += 1
if genome1[site] == ancestor[
site]: # counts up the number of shared sites that were inherited from the ancestor
a += 1
# s1_tree_location = scaled_tree_string.find(strain_names[s1])
# s2_tree_location = scaled_tree_string.find(strain_names[s2])
# start_length_1 = scaled_tree_string.find(':', s1_tree_location) + 1
# x1 = scaled_tree_string.find(',', start_length_1)
# y1 = scaled_tree_string.find(')', start_length_1)
# if x1 == -1:
# end_length_1 = y1
# elif y1 == -1:
# end_length_1 = x1
# else:
# end_length_1 = min(x1,y1)
# start_length_2 = scaled_tree_string.find(':', s2_tree_location) + 1
# x2 = scaled_tree_string.find(',', start_length_2)
# y2 = scaled_tree_string.find(')', start_length_2)
# if x2 == -1:
# end_length_2 = y2
# elif y2 == -1:
# end_length_2 = x2
# else:
# end_length_2 = min(x2,y2)
# length_1 = float(scaled_tree_string[start_length_1:end_length_1])
# length_2 = float(scaled_tree_string[start_length_2:end_length_2])
# MRCA = find_MRCA(strain1, strain2, parents) # the Most Recent Common Ancestor between the two strains
pair_distances = get_distances_to_MRCA(
strain1, strain2, MRCA, tree_string, strain_names, parents,
distances
) # gets the total lengths of the branches back to the MRCA of strain 1 and strain 2
distance_1 = pair_distances[
'distance_1'] # the distance from strain 1 to the MRCA
distance_2 = pair_distances[
'distance_2'] # the distance from strain 2 to the MRCA
m_1 = int(distance_1 * L +
1) # the number of mutations that occurred on strain 1
m_2 = int(distance_1 * L +
1) # the number of mutations that occurred on strain 2
generations_1 = int(
m_1 / mu
) # the number of generations over which these mutations occurred on strain 1
generations_2 = int(
m_2 / mu
) # the number of generations over which these mutations occurred on strain 2
c = expected_c_given_ms(
L, m_1, m_2, mu, generations_1, generations_2, kappa, 0.5
) # the expected number of convergent mutations between strain 1 and strain 2
# fills in the appropriate values to the S,C,A,R matrices for the current strain pair
SHARED[s1, s2] = s
SHARED[s2, s1] = s
CONVERGENT[s1, s2] = c
CONVERGENT[s2, s1] = c
ANCESTRAL[s1, s2] = a
ANCESTRAL[s2, s1] = a
RECOMBINANT[s1, s2] = s - a - c
RECOMBINANT[s2, s1] = s - a - c
count += 1
# print('\n\nCompleted strain pairing ' + str(count) + ' out of ' + str(n**2) + '\n\n')
# return {'strain_names': strain_names, 'Convergent': CONVERGENT}
return {
'strain_names': strain_names,
'Shared': SHARED,
'Convergent': CONVERGENT,
'Ancestral': ANCESTRAL,
'Recombinant': RECOMBINANT
}
# ancestral_tree_string = list(tree_file)[0]
# internal_nodes = get_internal_nodes(os.path.join(raxml_path, ancestral_alignment))
# internal_nodes,ancestral_tree_string = rename_ancestors(internal_nodes, strain_names, ancestral_tree_string)
# all_nodes = {}
# for key in strains.keys():
# all_nodes[key] = strains[key]
# for key in internal_nodes.keys():
# all_nodes[key] = internal_nodes[key]
# all_node_names = list(all_nodes.keys())
total_pairs = int(
(n * (n - 1)) /
2) # the total number of strain pairs that will be compared
pi = pi_value(strains)
theta = theta_value(
strains) # proportion of the genome that is polymorphic
# complete_tree_string = merge_trees(rooted_tree_string, ancestral_tree_string)
kappa_file = open(kappa_file, 'r')
kappa = float(list(kappa_file)[0])
min_m = get_min_m(
strains, L
) # minimum number of mutations that could account for all the polymorphisms in the species
max_m = get_max_m(strains, L, rooted_tree_string)
accurate_tree, trials = scale_branch_lengths(L, rooted_tree_string, min_m,
max_m, pi, theta, kappa, 1)
with open(output_file + '.csv', 'w', newline='') as f:
writer = csv.writer(f)
# runs the functions to get pi, theta, GC% average, and GC% standard deviation for each species and write them into a .csv file
# time complexity: O(n^4), where n is the length of the strains
path = 'C:/Users/Owner/Documents/UNCG REU/Project/concatenates/other' # path where the .fa files are located
with open(('species_params2.csv'), 'w', newline = '') as f:
writer = csv.writer(f)
writer.writerow(['Species', 'Genome Length', 'pi', 'theta', 'GC%']) # column headers
for filename in glob.glob(os.path.join(path, '*.fa')): # finds the values for each species
species = read_in_strains(filename)
name = filename[len(path)+1:len(filename)-3] # filename.strip('C:/Users/Owner/Documents/UNCG REU/Project/Recombination-Rates/concatenates').strip('/concat_') # strips off everything but the actual species name
# name = filename.strip('C:/Users/Owner/Documents/UNCG REU/Project/Recombination-Rates/concatenates').strip('/concat_') # strips off everything but the actual species name
print(name)
size = species_size(species)
length = genome_length(species)
pi = str(pi_value(species))
theta = str(theta_value(species))
# GC_comp = list(nucleotide_composition(species))
GC_average = nucleotide_composition(species)
# GC_average = GC_comp[0]
# GC_stdev = GC_comp[1]
writer.writerow([name, length, size, pi, theta, GC_average])
# print(filename)
# print('pi = ' + str(pi))
# print('theta = ' + str(theta))
# print('GC% = ' + str(GC_comp))
# print('\n')
print(name)
def get_SCAR_matrices(species_alignment, ancestral_alignment, kappa_file, mu,
species):
reduced_species_alignment = '/mnt/c/Users/Owner/Documents/UNCG/Project/standard-RAxML/done_species/' + species + 'concat_universal.fa.reduced'
raxml_path = '/mnt/c/Users/Owner/Documents/UNCG/Project/standard-RAxML/done_species/' + species # uberculosis'
tree_file = 'RAxML_bestTree.tree'
rooted_tree_file = 'RAxML_rootedTree.root'
# ancestral_alignment = 'RAxML_marginalAncestralStates.anc'
ancestral_tree_file = 'RAxML_nodeLabelledRootedTree.anc'
# get_tree_string(species_alignment, raxml_path)
reduced = os.path.exists(reduced_species_alignment)
# get_tree_root(tree_file, raxml_path)
# get_ancestors(rooted_tree_file, species_alignment, raxml_path, reduced)
if not reduced:
strains = read_in_strains(species_alignment)
else:
strains = read_in_reduced_strains(
reduced_species_alignment
) # dictionary with the genomes of all the strains; key = strain name, value = genome
# for strain in strains.keys():
# print(strain)
# print(strains[strain][:10])
L = genome_length(strains) # number of base pairs in the genome
n = species_size(strains) # number of extant strains
strain_names = list(strains.keys()) # list of all the extant strain names
SHARED = np.empty(
[n, n], dtype=np.float, order='C'
) # a matrix of the number of nucleotides shared between two strains; the (i,j) entry is the number of nucleotides that are the same between strain i and strain j
CONVERGENT = np.empty(
[n, n], dtype=np.float, order='C'
) # a matrix of the number of nucleotides that match due to convergent mutation between two strains; the (i,j) entry is the number of convergent mutations between strain i and strain j
ANCESTRAL = np.empty(
[n, n], dtype=np.float, order='C'
) # a matrix of the number of nucleotides that match due to direct inheritence from the ancestor; the (i,j) entry is the number of nucleotides that were inherited by both strain i and strain j
RECOMBINANT = np.empty(
[n, n], dtype=np.float, order='C'
) # a matrix of the number of nucleotides that match due to a recombination event; the (i,j) entry is the number of nucleotides that were recombined between strain i and strain j
RATES = np.empty([n, n], dtype=np.float, order='C')
tree_file = open((os.path.join(raxml_path, rooted_tree_file)), 'r')
rooted_tree_string = list(tree_file)[0]
tree_file = open((os.path.join(raxml_path, ancestral_tree_file)), 'r')
ancestral_tree_string = list(tree_file)[0]
# strains = read_in_strains(species_alignment) # dictionary with the genomes of all the strains; key = strain name, value = genome
internal_nodes = get_internal_nodes(
os.path.join(raxml_path, ancestral_alignment))
# print(internal_nodes.keys())
# all_nodes = internal_nodes
internal_nodes, ancestral_tree_string = rename_ancestors(
internal_nodes, strain_names, ancestral_tree_string)
all_nodes = {}
for key in strains.keys():
all_nodes[key] = strains[key]
for key in internal_nodes.keys():
all_nodes[key] = internal_nodes[key]
# strain_names = list(strains.keys()) # list of all the extant strain names
all_node_names = list(all_nodes.keys())
# print(strain_names)
print(all_node_names)
# n = species_size(strains) # number of extant strains
total_pairs = int(
(n * (n - 1)) /
2) # the total number of strain pairs that will be compared
# L = genome_length(strains) # number of base pairs in the genome
pi = pi_value(strains)
theta = theta_value(
strains) # proportion of the genome that is polymorphic
# print(theta)
# print(n)
# mu = (theta)/(2*n) # mutation rate in mutations per base pair per generation
# print(mu)
# tree_file = open((os.path.join(raxml_path, rooted_tree_file)), 'r')
# rooted_tree_string = list(tree_file)[0]
# tree_file = open((os.path.join(raxml_path, ancestral_tree_file)), 'r')
# ancestral_tree_string = list(tree_file)[0]
# print(rooted_tree_string)
# print(ancestral_tree_string)
complete_tree_string = merge_trees(rooted_tree_string,
ancestral_tree_string)
print(complete_tree_string)
kappa_file = open(kappa_file, 'r')
kappa = float(list(kappa_file)[0])
# min_m = get_min_m(strains, L) # minimum number of mutations that could account for all the polymorphisms in the species
# max_m = get_max_m(strains, L, complete_tree_string)
#^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
###############################################################################
##### CHANGE THIS!!!!!!!!!!!!!!!!!!!!! ########################################
###############################################################################
# scaled_tree_string = complete_tree_string
scaled_tree_string = scale_newick_format_tree(complete_tree_string)
# scaled_tree_string = scale_branch_lengths(L, complete_tree_string, min_m, max_m, pi, theta, kappa, 1) # scale_newick_format_tree(strains, L, min_m, tree_string, 0) # the tree_string scaled by min_m
# L, tree_string, min_m, max_m, real_pi, real_theta, kappa
phylogeny = pyvolve.read_tree(tree=scaled_tree_string)
# pyvolve.print_tree(phylogeny)
g = open('scaled_tree.txt', 'w')
g.write(scaled_tree_string)
g.close()
# updated_tree_info = name_nodes(tree_string, strain_names)
# scaled_tree_string = updated_tree_info['tree_string'] # version of the tree_string where every node is labeled
# new_nodes = updated_tree_info['new_nodes'] # the new node names that were added
# all_nodes = all_node_names + new_nodes # list of all the node names in the pyhlogenetic tree
# print(all_nodes)
# parents = find_parents(strain_names, tree_string) # a dictionary of the sequence of parents of each strain; key = strain name, value = list of the parents in order of increasing distance from the strain
parents = find_parents(all_node_names, scaled_tree_string)
# print('found parents')
# print(parents)
distances = get_branch_lengths(
all_node_names, scaled_tree_string
) # a dictionary of the distances of each strain to its closest ancestor; key = strain name, value = distance to its closest ancestor
# print('found distances')
# print(distances)
count = 1 # a counter for the current strain pair number that is being processed
total = 0
for s1 in range(n): # allows each strain to be strain 1
strain1 = strain_names[s1]
genome1 = strains[strain1]
SHARED[s1, s1] = L
CONVERGENT[
s1,
s1] = 0 # there can be no convergent mutations between a strain and itself
ANCESTRAL[s1, s1] = L
RECOMBINANT[s1, s1] = 0
for s2 in range(s1 + 1,
n): # allows each strain after strain 1 to be strain 2
strain2 = strain_names[s2]
genome2 = strains[strain2]
MRCA = find_MRCA(
strain1, strain2, parents
) # the Most Recent Common Ancestor between the two strains
MRCA_genome = all_nodes[MRCA]
s, a = get_s_a(genome1, genome2, MRCA_genome, L)
# print('got s and a')
c, pair_distances = get_c(strain1, strain2, MRCA, parents,
scaled_tree_string, distances, L, mu,
kappa)
# print('got c')
r = s - c - a
# print(pair_distances['distance_1'])
# print(pair_distances['distance_2'])
# print(L)
# fills in the appropriate values to the S,C,A,R matrices for the current strain pair
SHARED[s1, s2] = s
SHARED[s2, s1] = s
CONVERGENT[s1, s2] = c
CONVERGENT[s2, s1] = c
ANCESTRAL[s1, s2] = a
ANCESTRAL[s2, s1] = a
RECOMBINANT[s1, s2] = r
RECOMBINANT[s2, s1] = r
RATES[s1, s2] = r / int(pair_distances['distance_1'] * L + 1)
RATES[s2, s1] = r / int(pair_distances['distance_2'] * L + 1)
total += r / int(pair_distances['distance_1'] * L + 1)
total += r / int(pair_distances['distance_2'] * L + 1)
# print('\n\nCompleted strain pairing ' + str(count) + ' out of ' + str(total_pairs) + '\n\n')
count += 1
average_rate = total / total_pairs
print('The average recombination rate is ' + str(average_rate))
return {
'strain_names': strain_names,
'Shared': SHARED,
'Convergent': CONVERGENT,
'Ancestral': ANCESTRAL,
'Recombinant': RECOMBINANT,
'Rates': RATES,
'average': average_rate
}