},
'DM': {
'name': 'Insect',
}
}
for celltype in celltypes:
print("Loading celltype: {celltype:s}".format(celltype=celltype))
#
rGfile = '../../04-network/results/network/{celltype:s}/net-{celltype:s}-{network:s}-{threshold:s}.gpickle'.format(
celltype=celltype, network=network, threshold=threshold_str)
G = nx.read_gpickle(rGfile)
#
for layer in layers:
print('Separate layer {layer:s}'.format(layer=layer))
Gl = get_network_layer(G, layer)
#
data[layer][celltype] = {'graph': Gl}
# Compute Jaccard
r = []
for celltype_i, celltype_j in combinations(celltypes, 2):
for layer in layers:
G_i = data[layer][celltype_i]['graph']
G_j = data[layer][celltype_j]['graph']
genes_i = G_i.nodes()
genes_j = G_j.nodes()
# For "sign indeterminacy"
np.random.seed(1)
#
# Load Network
#
print('Reading {celltype:s}-{network:s}-{threshold:s} Network'.format(
celltype=celltype, network=network, threshold=threshold_str))
rGfile_gpickle = 'results/network/{celltype:s}/net-{celltype:s}-{network:s}-{threshold:s}.gpickle'.format(
celltype=celltype, network=network, threshold=threshold_str)
G = nx.read_gpickle(rGfile_gpickle)
# SVD per Layer
for layer in ['HS', 'MM', 'DM']:
print('Isolate {layer:s} Layer'.format(layer=layer))
Gt = get_network_layer(G, layer=layer)
#
dfG = pd.DataFrame(data={
'gene': [d.get('label', None) for n, d in Gt.nodes(data=True)]
},
index=Gt.nodes)
#
print('Extract Adjacency Matrix')
M = nx.to_numpy_matrix(Gt)
print('Calculating PCA (sklearn)')
pca = PCA(n_components=None, svd_solver='full')
res = pca.fit(M).transform(M)
#
columns = ['{:d}c'.format(i) for i in range(1, components + 1)]
df_pca = pd.DataFrame(res[:, 0:components],
args = parser.parse_args()
celltype = args.celltype # spermatocyte or enterocyte
network = 'full'
attribute = 'combined_score'
threshold = args.threshold
print('Loading Full Network')
path_net = '../../04-network/results/network/{celltype:s}/'.format(
celltype=celltype)
rGfile_gpickle = path_net + 'net-{celltype:s}-{network:s}.gpickle'.format(
celltype=celltype, network=network)
G = nx.read_gpickle(rGfile_gpickle)
print('Separate Layers')
HSG = get_network_layer(G, 'HS')
MMG = get_network_layer(G, 'MM')
DMG = get_network_layer(G, 'DM')
print('Get edge weights')
values_HS = sorted([
d[attribute] / 1000
for i, j, d in HSG.edges(data=True) if attribute in d
],
reverse=True)
values_MM = sorted([
d[attribute] / 1000
for i, j, d in MMG.edges(data=True) if attribute in d
],
reverse=True)
values_DM = sorted([
from utils import get_network_layer, ensurePathExists
if __name__ == '__main__':
threshold = 0.5
threshold_str = str(threshold).replace('.', 'p')
path_net = '../../04-network/results/network/{celltype:s}/'.format(celltype='spermatocyte')
rG_spermatocyte_file_gpickle = path_net + 'net-{celltype:s}-{network:s}-{threshold:s}.gpickle'.format(celltype='spermatocyte', network='conserved', threshold=threshold_str)
print('Load Spermatocyte graph')
Gs = nx.read_gpickle(rG_spermatocyte_file_gpickle)
print('Separate Layers')
HSGs = get_network_layer(Gs, 'HS')
MMGs = get_network_layer(Gs, 'MM')
DMGs = get_network_layer(Gs, 'DM')
path_net = '../../04-network/results/network/{celltype:s}/'.format(celltype='enterocyte')
rG_enterocyte_file_gpickle = path_net + 'net-{celltype:s}-{network:s}-{threshold:s}.gpickle'.format(celltype='enterocyte', network='conserved', threshold=threshold_str)
print('Load Enterocyte graph')
Ge = nx.read_gpickle(rG_enterocyte_file_gpickle)
print('Separate Layers')
HSGe = get_network_layer(Ge, 'HS')
MMGe = get_network_layer(Ge, 'MM')
DMGe = get_network_layer(Ge, 'DM')
dict_data = {
'HS': {
celltype = 'spermatocyte'
threshold = 0.5
threshold_str = str(threshold).replace('.', 'p')
network = 'full'
path_net = '../../04-network/results/network/{celltype:s}/'.format(
celltype=celltype)
rG_full_file_gpickle = path_net + 'net-{celltype:s}-{network:s}.gpickle'.format(
celltype=celltype, network=network)
print('Load {celltype:s} {network:s} graph'.format(celltype=celltype,
network=network))
Gf = nx.read_gpickle(rG_full_file_gpickle)
print('Separate Layers')
HSGf = get_network_layer(Gf, 'HS')
MMGf = get_network_layer(Gf, 'MM')
DMGf = get_network_layer(Gf, 'DM')
network = 'conserved'
path_net = '../../04-network/results/network/{celltype:s}/'.format(
celltype=celltype)
rG_con_file_gpickle = path_net + 'net-{celltype:s}-{network:s}-{threshold:s}.gpickle'.format(
celltype=celltype, network=network, threshold=threshold_str)
print('Load {celltype:s} {network:s} graph'.format(celltype=celltype,
network=network))
Gc = nx.read_gpickle(rG_con_file_gpickle)
print('Separate Layers')
HSGc = get_network_layer(Gc, 'HS')
MMGc = get_network_layer(Gc, 'MM')
celltype = 'enterocyte'
threshold = 0.5
threshold_str = str(threshold).replace('.', 'p')
network = 'thr'
path_net = '../../04-network/results/network/{celltype:s}/'.format(
celltype=celltype)
rG_thr_file_gpickle = path_net + 'net-{celltype:s}-{network:s}-{threshold:s}.gpickle'.format(
celltype=celltype, network=network, threshold=threshold_str)
print('Load {celltype:s} {network:s} graph'.format(celltype=celltype,
network=network))
Gt = nx.read_gpickle(rG_thr_file_gpickle)
print('Separate Layers')
HSGt = get_network_layer(Gt, 'HS')
MMGt = get_network_layer(Gt, 'MM')
DMGt = get_network_layer(Gt, 'DM')
network = 'conserved'
path_net = '../../04-network/results/network/{celltype:s}/'.format(
celltype=celltype)
rG_con_file_gpickle = path_net + 'net-{celltype:s}-{network:s}-{threshold:s}.gpickle'.format(
celltype=celltype, network=network, threshold=threshold_str)
print('Load {celltype:s} {network:s} graph'.format(celltype=celltype,
network=network))
Gc = nx.read_gpickle(rG_con_file_gpickle)
print('Separate Layers')
HSGc = get_network_layer(Gc, 'HS')
MMGc = get_network_layer(Gc, 'MM')
default='spermatocyte',
type=str,
choices=['spermatocyte', 'enterocyte'],
help=
"Cell type. Must be either 'spermatocyte' or 'enterocyte'. Defaults to spermatocyte"
)
args = parser.parse_args()
celltype = args.celltype # spermatocyte or enterocyte
print('Reading Network')
rGfile_gpickle = 'results/net-{celltype:s}.gpickle'.format(
celltype=celltype)
G = nx.read_gpickle(rGfile_gpickle)
DMG = get_network_layer(G, 'DM')
core_DM = nx.get_node_attributes(DMG, name='core')
gene_DM = nx.get_node_attributes(DMG, name='label')
df_DM_m = pd.DataFrame(data={'gene': gene_DM, 'core': core_DM})
df_DM_m['core'] = df_DM_m['core'].fillna(False)
r = []
Gt = DMG.copy()
r.append([
None,
Gt.number_of_nodes(),
Gt.number_of_edges(),
len([
i for i, d in Gt.nodes(data=True) if d.get('core', False) == True
])
data = {
'HS': {},
'MM': {},
'DM': {}
}
print('-- Conserved --')
for celltype in celltypes:
print('Loading {celltype:s} {network:s} {threshold:s}'.format(celltype=celltype, network='conserved', threshold=threshold_str))
path_net = '../../04-network/results/network/{celltype:s}/'.format(celltype=celltype)
rGc_file_gpickle = path_net + 'net-{celltype:s}-{network:s}-{threshold:s}.gpickle'.format(celltype=celltype, network='conserved', threshold=threshold_str)
Gc = nx.read_gpickle(rGc_file_gpickle)
for layer in layers:
print('Separating layer {layer:s}'.format(layer=layer))
Gcl = get_network_layer(Gc, layer)
conserved_genes = set(Gcl.nodes())
data[layer][celltype] = conserved_genes
for layer in layers:
print('Calculating venn {layer:s} conserved'.format(layer=layer))
ns = data[layer]
for intersected, unioned, count in venn_count(ns):
print('|{}{}| = {}'.format(' & '.join(sorted(intersected)), ' - ' + ' - '.join(sorted(unioned)) if unioned else '', count))
#
# Non-conserved
#