def run_in_out_entropy_comparision(dataset, graphml_file, folder):
g = util.read_graphml(graphml_file)
in_residues = get_residues_from_graph(g)
avg_entropy_dict = {}
avg_in_nx_entropy_dict = {}
avg_out_nx_entropy_dict = {}
avg_entropy_dict['dataset'] = dataset
avg_in_nx_entropy_dict['dataset'] = dataset
avg_out_nx_entropy_dict['dataset'] = dataset
proteins = ['ha', 'na', 'm1', 'm2', 'np', 'pb1', 'pb2', 'pa', 'ns1', 'ns2']
for protein in proteins:
file = folder + os.sep + protein + '.afasta'
alignment = AlignIO.read(file, 'fasta')
sequences = [x.seq for x in alignment]
entropies = util.entropy_all_positions(sequences)
entropy_avg, in_network_entropy_avg, out_network_entropy_avg = in_out_entropies(
entropies, in_residues[protein], folder, protein)
avg_entropy_dict[protein] = entropy_avg
avg_in_nx_entropy_dict[protein] = in_network_entropy_avg
avg_out_nx_entropy_dict[protein] = out_network_entropy_avg
print('averages for ', dataset, ' ', protein, ' ',
in_network_entropy_avg, out_network_entropy_avg)
return avg_entropy_dict, avg_in_nx_entropy_dict, avg_out_nx_entropy_dict
def get_cooccurences(graphml_file, fasta_folder):
top_50_edges_cooccurence_counts = []
bottom_50_edges_cooccurence_counts = []
g = util.read_graphml(graphml_file)
top_n_edges = util.get_edges(g, start=0, end=50, reverse=True)
bottom_n_edges = util.get_edges(g, start=0, end=50, reverse=False)
for edge in top_n_edges:
residue1 = int(edge[0].split('_')[1]) - 1
protein1 = edge[0].split('_')[0]
residue2 = int(edge[1].split('_')[1]) - 1
protein2 = edge[1].split('_')[0]
file1 = fasta_folder + os.sep + protein1 + '.afasta'
file2 = fasta_folder + os.sep + protein2 + '.afasta'
rcc = perform_residue_analysis(file1, file2, residue1, residue2)
print(edge)
print(sorted(rcc.items(), key=operator.itemgetter(1), reverse=True))
top_50_edges_cooccurence_counts.append(rcc)
print('=====================')
for edge in bottom_n_edges:
residue1 = int(edge[0].split('_')[1]) - 1
protein1 = edge[0].split('_')[0]
residue2 = int(edge[1].split('_')[1]) - 1
protein2 = edge[1].split('_')[0]
file1 = fasta_folder + os.sep + protein1 + '.afasta'
file2 = fasta_folder + os.sep + protein2 + '.afasta'
rcc = perform_residue_analysis(file1, file2, residue1, residue2)
print(sorted(rcc.items(), key=operator.itemgetter(1), reverse=True))
bottom_50_edges_cooccurence_counts.append(rcc)
return top_50_edges_cooccurence_counts, bottom_50_edges_cooccurence_counts
def run_in_out_acc_comparision(dataset, graphml_file, folder):
avg_in_acc_dict = {}
avg_out_acc_dict = {}
avg_in_acc_dict['dataset'] = dataset
avg_out_acc_dict['dataset'] = dataset
g = util.read_graphml(graphml_file)
in_residues = get_residues_from_graph(g)
dssp = DSSPData()
avg_in, avg_out = run_in_out_acc_ha(in_residues, dssp, g, folder)
avg_in_acc_dict['ha'] = avg_in
avg_out_acc_dict['ha'] = avg_out
avg_in, avg_out = run_in_out_acc_m1(in_residues, dssp, g, folder)
avg_in_acc_dict['m1'] = avg_in
avg_out_acc_dict['m1'] = avg_out
avg_in, avg_out = run_in_out_acc_na(in_residues, dssp, g, folder)
avg_in_acc_dict['na'] = avg_in
avg_out_acc_dict['na'] = avg_out
avg_in, avg_out = run_in_out_acc_np(in_residues, dssp, g, folder)
avg_in_acc_dict['np'] = avg_in
avg_out_acc_dict['np'] = avg_out
avg_in, avg_out = run_in_out_acc_ns1(in_residues, dssp, g, folder)
avg_in_acc_dict['ns1'] = avg_in
avg_out_acc_dict['ns1'] = avg_out
return avg_in_acc_dict, avg_out_acc_dict
def run_in_out_acc_comparision(dataset, graphml_file, folder):
avg_in_acc_dict = {}
avg_out_acc_dict = {}
avg_in_acc_dict['dataset'] = dataset
avg_out_acc_dict['dataset'] = dataset
g = util.read_graphml(graphml_file)
in_residues = get_residues_from_graph(g)
dssp = DSSPData()
avg_in, avg_out = run_in_out_acc_ha(in_residues, dssp, g, folder)
avg_in_acc_dict['ha'] = avg_in
avg_out_acc_dict['ha'] = avg_out
avg_in, avg_out = run_in_out_acc_m1(in_residues, dssp, g, folder)
avg_in_acc_dict['m1'] = avg_in
avg_out_acc_dict['m1'] = avg_out
avg_in, avg_out = run_in_out_acc_na(in_residues, dssp, g, folder)
avg_in_acc_dict['na'] = avg_in
avg_out_acc_dict['na'] = avg_out
avg_in, avg_out = run_in_out_acc_np(in_residues, dssp, g, folder)
avg_in_acc_dict['np'] = avg_in
avg_out_acc_dict['np'] = avg_out
avg_in, avg_out = run_in_out_acc_ns1(in_residues, dssp, g, folder)
avg_in_acc_dict['ns1'] = avg_in
avg_out_acc_dict['ns1'] = avg_out
return avg_in_acc_dict, avg_out_acc_dict
def get_cooccurences(graphml_file, fasta_folder):
top_50_edges_cooccurence_counts = []
bottom_50_edges_cooccurence_counts = []
g = util.read_graphml(graphml_file)
top_n_edges = util.get_edges(g, start=0, end=50, reverse=True)
bottom_n_edges = util.get_edges(g, start=0, end=50, reverse=False)
for edge in top_n_edges:
residue1 = int(edge[0].split('_')[1])-1
protein1 = edge[0].split('_')[0]
residue2 = int(edge[1].split('_')[1])-1
protein2 = edge[1].split('_')[0]
file1 = fasta_folder + os.sep + protein1 + '.afasta'
file2 = fasta_folder + os.sep + protein2 + '.afasta'
rcc = perform_residue_analysis(file1, file2, residue1, residue2)
print(edge)
print(sorted(rcc.items(), key=operator.itemgetter(1), reverse=True))
top_50_edges_cooccurence_counts.append(rcc)
print('=====================')
for edge in bottom_n_edges:
residue1 = int(edge[0].split('_')[1])-1
protein1 = edge[0].split('_')[0]
residue2 = int(edge[1].split('_')[1])-1
protein2 = edge[1].split('_')[0]
file1 = fasta_folder + os.sep + protein1 + '.afasta'
file2 = fasta_folder + os.sep + protein2 + '.afasta'
rcc = perform_residue_analysis(file1, file2, residue1, residue2)
print(sorted(rcc.items(), key=operator.itemgetter(1), reverse=True))
bottom_50_edges_cooccurence_counts.append(rcc)
return top_50_edges_cooccurence_counts, bottom_50_edges_cooccurence_counts
def run(infilename):
ingraph = read_graphml(infilename)
outgraph = create_protein_graph(ingraph)
write_graphml(outgraph, infilename)
plot_file = infilename.split('.')[0] + '.png'
create_plot(outgraph, plot_file)
#run('u800d_05_B01b.graphml')
#compare_two_graphs('all_05_01.graphml', '800d_05_01.graphml')
def compare_two_graphs(file1, file2):
g1 = read_graphml(file1)
g2 = read_graphml(file2)
nodes1 = g1.nodes()
nodes2 = g2.nodes()
matches = []
in1only = []
in2only = []
for node in nodes1:
if node in nodes2:
matches.append(node)
else:
in1only.append(node)
for node in nodes2:
if node not in nodes1:
in2only.append(node)
print(len(nodes1), len(nodes2), len(matches), len(in1only), len(in2only))
def run_in_out_acc_comparision():
graphml_file = sys.argv[1]
folder = sys.argv[2]
g = util.read_graphml(graphml_file)
in_residues = get_residues_from_graph(g)
dssp = DSSPData()
run_in_out_acc_ha(in_residues, dssp, g, folder)
run_in_out_acc_m1(in_residues, dssp, g, folder)
run_in_out_acc_na(in_residues, dssp, g, folder)
run_in_out_acc_np(in_residues, dssp, g, folder)
run_in_out_acc_ns1(in_residues, dssp, g, folder)
def run(folder, infilename, title):
ingraph = read_graphml(folder + os.sep + infilename)
try:
#avg_cluster_coeff = nx.average_clustering(ingraph)
#print('average clustering for ' + title + " = " + str(avg_cluster_coeff))
avg_deg_coeff = nx.average_degree_connectivity(ingraph)
print('average degree for ' + title + ' = ' + str(avg_deg_coeff))
except Exception as e:
print (e.__str__())
'''
def run_in_out_entropy_comparision():
graphml_file = sys.argv[1]
folder = sys.argv[2]
g = util.read_graphml(graphml_file)
in_residues = get_residues_from_graph(g)
proteins = ['ha', 'na', 'm1', 'm2', 'np', 'pb1', 'pb2', 'pa', 'ns1', 'ns2']
for protein in proteins:
file = folder + os.sep + protein + '.afasta'
alignment = AlignIO.read(file, 'fasta')
sequences = [x.seq for x in alignment]
entropies = util.entropy_all_positions(sequences)
in_network_entropy_avg, out_network_entropy_avg = in_out_entropies(entropies, in_residues[protein], folder, protein)
print('averages for ', protein, ' ', in_network_entropy_avg, out_network_entropy_avg)
def run_pairwise_comparision():
graphml_file = sys.argv[1]
folder = sys.argv[2]
g = util.read_graphml(graphml_file)
top_ten_edges = get_best_edges_from_graph(g, 10)
for edge in top_ten_edges:
residue1 = int(edge[0].split('_')[1])-1
protein1 = edge[0].split('_')[0]
residue2 = int(edge[1].split('_')[1])-1
protein2 = edge[1].split('_')[0]
print(protein1, protein2, residue1, residue2)
file1 = folder + os.sep + protein1 + '.afasta'
file2 = folder + os.sep + protein2 + '.afasta'
perform_residue_analysis(file1, file2, residue1, residue2)
def run_in_out_entropy_comparision(dataset, graphml_file, folder):
g = util.read_graphml(graphml_file)
in_residues = get_residues_from_graph(g)
avg_entropy_dict = {}
avg_in_nx_entropy_dict = {}
avg_out_nx_entropy_dict = {}
avg_entropy_dict['dataset'] = dataset
avg_in_nx_entropy_dict['dataset'] = dataset
avg_out_nx_entropy_dict['dataset'] = dataset
proteins = ['ha', 'na', 'm1', 'm2', 'np', 'pb1', 'pb2', 'pa', 'ns1', 'ns2']
for protein in proteins:
file = folder + os.sep + protein + '.afasta'
alignment = AlignIO.read(file, 'fasta')
sequences = [x.seq for x in alignment]
entropies = util.entropy_all_positions(sequences)
entropy_avg, in_network_entropy_avg, out_network_entropy_avg = in_out_entropies(entropies, in_residues[protein], folder, protein)
avg_entropy_dict[protein] = entropy_avg
avg_in_nx_entropy_dict[protein] = in_network_entropy_avg
avg_out_nx_entropy_dict[protein] = out_network_entropy_avg
print('averages for ', dataset, ' ', protein, ' ', in_network_entropy_avg, out_network_entropy_avg)
return avg_entropy_dict, avg_in_nx_entropy_dict, avg_out_nx_entropy_dict
def run(folder, infilename, title):
ingraph = read_graphml(folder + '\\' + infilename)
create_clustering_plot(ingraph, folder=folder, title=title)
create_degree_plot(ingraph, folder=folder, title=title)
def run(folder, infilename, title):
ingraph = read_graphml(folder + os.sep + infilename)
outgraph = create_protein_graph(ingraph)
write_macro_graphml(outgraph, folder, infilename)
plot_file = infilename.split('.')[0] + '.png'
create_plot(outgraph, folder, plot_file, title)
