def construct_trn(path):
LOGGER.info("{0:*^78s}".format("Construct TRN"))
version = os.path.basename(path)
if not version:
version = os.path.basename(os.path.dirname(path))
LOGGER.info("{0:*^78s}".format(version))
# load objects so that they are in memory
pyorganism.read_pickle(os.path.join(path, "conformations.pkl"))
t_units = pyorganism.read_pickle(os.path.join(path, "transcription_units.pkl"))
interactions = pyorganism.read_pickle(os.path.join(path, "interactions.pkl"))
assert all(pyreg.Conformation.has_key(triple[0], version) for triple in interactions),\
"unknown conformation in regulatory interactions"
assert all(pyreg.Promoter.has_key(triple[1], version) for triple in interactions),\
"unknown promoter in regulatory interactions"
# conformation-promoter directed regulatory network
cpn = nx.MultiDiGraph()
for (u, v, inter) in interactions:
first = pyreg.Conformation[u, version]
second = pyreg.Promoter[v, version]
cpn.add_edge(first, second, key=inter)
# TRN from CPN
trn = pyreg.TRN()
node_converter = NodeConverter(t_units)
for (u, v, k, d) in cpn.edges_iter(keys=True, data=True):
first = u.t_factor
for nbr in node_converter(v.unique_id):
trn.add_edge(first, nbr, key=k, **d.copy())
LOGGER.info("%d Nodes", len(trn))
LOGGER.info("%d Links", trn.size())
components = list(nx.connected_components(trn.to_undirected()))
LOGGER.info("%d Components", len(components))
LOGGER.info("Largest Component: %d", len(components[0]))
if len(components) > 1:
LOGGER.info("Second Largest Component: %d", len(components[1]))
pyorganism.write_pickle(trn, os.path.join(path, "trn.pkl"))
def compile_feature2gene(objects_path):
LOGGER.info("{0:*^78s}".format("Compile Gene Feature Map"))
version = os.path.basename(objects_path)
if not version:
version = os.path.basename(os.path.dirname(objects_path))
global VERSION
VERSION = version
LOGGER.info("{0:*^78s}".format(version))
names = compile_names(CONFIG["data_paths"], CONFIG["data_load"])
LOGGER.info("Loading genes")
genes = pyorganism.read_pickle(os.path.join(objects_path, "genes.pkl"))
LOGGER.info("Finding gene names")
finder = pyorganism.FindObject(genes,
targets=[gene.name.lower() for gene in genes])
(feature2gene, gap) = compile_feature_map(genes, names, finder)
synonyms = synonym_finder(genes)
LOGGER.info("Finding gene names by synonyms")
gap = extend_feature_map(feature2gene, gap, synonyms)
LOGGER.info("Missing %d gene names", len(gap))
LOGGER.info("Fuzzy search of gene names (threshold %d%%)", CONFIG["threshold"])
LOGGER.setLevel(logging.DEBUG)
gap = fuzzy_extension(feature2gene, gap, finder, CONFIG["threshold"])
LOGGER.info("Fuzzy search of gene names by synonyms (threshold %d%%)", CONFIG["threshold"])
gap = fuzzy_extension(feature2gene, gap, synonyms, CONFIG["threshold"])
LOGGER.info("Manual update")
manual_name_updates(feature2gene)
num = sum(1 for gene in feature2gene.itervalues() if gene)
LOGGER.info("Final map contains %d names and %d genes (%3.2f%%)", len(feature2gene),
num, 100.0 * num / len(feature2gene))
pyorganism.write_pickle(feature2gene, os.path.join(objects_path, "feature2gene.pkl"))
def update_map(map_path, frame_path, update_from, version=""):
LOGGER.info("{0:*^78s}".format("Update Gene IDs"))
map_name = os.path.splitext(os.path.basename(map_path))[0]
LOGGER.info("{0:*^78s}".format(map_name))
base_path = os.path.dirname(map_path)
if not version:
version = os.path.basename(base_path)
LOGGER.info("{0:*^78s}".format(version))
# load into memory
LOGGER.info("Loading genes")
pyorganism.read_pickle(os.path.join(base_path, "genes.pkl"))
LOGGER.info("Reading data frame")
hdf_key = "/%s" % (map_name,)
mapping = pandas.read_hdf(frame_path, hdf_key)
# frame_info(mapping)
id2gene = dict()
for row in mapping[[version, update_from]].itertuples():
if not row[1]:
id2gene[row[0]] = pyreg.Gene.get(row[2], None, version)
elif row[1] != row[2]:
id2gene[row[0]] = pyreg.Gene.get(row[2], None, version)
else:
id2gene[row[0]] = pyreg.Gene.get(row[1], None, version)
pyorganism.write_pickle(id2gene, os.path.join(base_path, "corrected_" +
map_name + ".pkl"))
def compile_name2gene(objects_path):
LOGGER.info("{0:*^78s}".format("Compile Gene Name Map"))
full_data = compile_names(CONFIG["data_paths"], CONFIG["data_load"])
version = os.path.basename(objects_path)
if not version:
version = os.path.basename(os.path.dirname(objects_path))
global VERSION
VERSION = version
LOGGER.info("{0:*^78s}".format(version))
LOGGER.info("Loading genes")
genes = pyorganism.read_pickle(os.path.join(objects_path, "genes.pkl"))
LOGGER.info("Finding gene names")
names = set(full_data["name"].unique())
if numpy.nan in names:
names.remove(numpy.nan)
names = sorted(names)
finder = pyorganism.FindObject(genes, "name")
(name2gene, name_gap) = compile_feature_map(genes, names, finder)
synonyms = synonym_finder(genes)
LOGGER.info("Finding gene names by synonyms")
name_gap = extend_feature_map(name2gene, name_gap, synonyms)
LOGGER.info("Missing %d gene names", len(name_gap))
LOGGER.info("Fuzzy search of gene names (threshold %d%%)", CONFIG["threshold"])
name_gap = fuzzy_extension(name2gene, name_gap, finder, CONFIG["threshold"])
# LOGGER.info("Fuzzy search of gene names by synonyms (threshold %d%%)", CONFIG["threshold"])
# name_gap = fuzzy_extension(name2gene, name_gap, synonyms, CONFIG["threshold"])
manual_name_updates(name2gene)
num = sum(1 for gene in name2gene.itervalues() if gene)
LOGGER.info("Final map contains %d names and %d genes (%3.2f%%)", len(name2gene),
num, 100.0 * num / len(name2gene))
LOGGER.info("Finding gene blattner numbers")
bnumbers = set(full_data["blattner"].unique())
if numpy.nan in bnumbers:
bnumbers.remove(numpy.nan)
bnumbers = sorted(bnumbers)
gene_finder = pyorganism.FindObject(genes, "bnumber")
(blattner2gene, blattner_gap) = compile_feature_map(genes, bnumbers, gene_finder)
LOGGER.info("Finding gene blattner numbers by synonyms")
blattner_gap = extend_feature_map(blattner2gene, blattner_gap, synonyms)
LOGGER.info("Missing %d gene blattner numbers", len(blattner_gap))
LOGGER.info("Fuzzy search of gene blattner numbers (threshold %d%%)", CONFIG["threshold"])
blattner_gap = fuzzy_extension(blattner2gene, blattner_gap, finder, CONFIG["threshold"])
# LOGGER.info("Fuzzy search of gene blattner numbers by synonyms (threshold %d%%)", CONFIG["threshold"])
# blattner_gap = fuzzy_extension(blattner2gene, blattner_gap, synonyms, CONFIG["threshold"])
num = sum(1 for gene in blattner2gene.itervalues() if gene)
LOGGER.info("Final map contains %d blattner numbers and %d genes (%3.2f%%)",
len(blattner2gene), num, 100.0 * num / len(blattner2gene))
verify_union(name2gene, blattner2gene, full_data)
pyorganism.write_pickle(name2gene, os.path.join(objects_path, "name2gene.pkl"))
pyorganism.write_pickle(blattner2gene, os.path.join(objects_path, "blattner2gene.pkl"))
def main_random(args):
locations = sorted(glob(os.path.join(args.in_path, args.glob)))
locations = [os.path.abspath(loc) for loc in locations]
pool = multiprocessing.Pool(args.nproc)
bar = ProgressBar(maxval=args.rnd_num, widgets=[Timer(), " ",
SimpleProgress(), " ", Percentage(), " ", Bar(), " ", ETA()])
rewire_name = "grn_rewired_{0:.1f}.pkl".format(args.prob)
switch_name = "grn_switched_{0:d}.pkl".format(args.flip_num)
for path in locations:
filename = os.path.join(path, "trn.pkl")
if not os.path.exists(filename):
continue
ver = version_from_path(path)
base_path = os.path.join(args.out_path, ver)
if not os.path.isdir(base_path):
os.makedirs(base_path)
LOGGER.info(ver)
trn = pyorg.read_pickle(filename)
# we consider null-models on the projected level since that is the level
# on which we evaluate topological quantities
net = pyreg.to_simple(trn.to_grn())
if args.run_rewire:
LOGGER.info("Rewiring with probability %.1f", args.prob)
tasks = [(net, args.prob)] * args.rnd_num
res_it = pool.imap_unordered(rewire, tasks)
rands = list()
bar.start()
for rnd in res_it:
rands.append(rnd)
bar += 1
bar.finish()
pyorg.write_pickle(rands, os.path.join(base_path, rewire_name))
if args.run_switch:
LOGGER.info("Switch-randomizing each edge %d times", args.flip_num)
tasks = [(net, args.flip_num)] * args.rnd_num
res_it = pool.imap_unordered(switch, tasks)
success = list()
rands = list()
bar.start()
for (rnd, rate) in res_it:
rands.append(rnd)
success.append(rate)
bar += 1
bar.finish()
pyorg.write_pickle(rands, os.path.join(base_path, switch_name))
LOGGER.info("mean flip success rate: %.3G +/- %.3G", np.mean(success),
np.std(success))
pool.close()
def rewiring(lb_view, versions, args):
bar = ProgressBar(maxval=args.rnd_num, widgets=[Timer(), " ",
SimpleProgress(), " ", Percentage(), " ", Bar(), " ", ETA()])
rands = list()
for ver in versions:
LOGGER.info(ver)
res_it = lb_view.map(rewire, [ver] * args.rnd_num, block=False, ordered=False)
bar.start()
for rng in res_it:
rands.append(rng)
bar += 1
bar.finish()
pyorg.write_pickle(rands, os.path.join(args.out_path, ver,
"trn_rewired_{0:.1f}.pkl".format(args.prob)))
lb_view.purge_results("all")
del rands[:] # activate garbage collection in loop
def randomisation(lb_view, versions, args):
bar = ProgressBar(maxval=args.rnd_num, widgets=[Timer(), " ",
SimpleProgress(), " ", Percentage(), " ", Bar(), " ", ETA()])
for ver in versions:
LOGGER.info(ver)
res_it = lb_view.map(null_model, [ver] * args.rnd_num, block=False, ordered=False)
bar.start()
rands = list()
success = list()
for (rnd_net, flip_rate) in res_it:
rands.append(rnd_net)
success.append(flip_rate)
bar +=1
bar.finish()
lb_view.purge_results("all")
LOGGER.info("mean flip success rate: %.3G +/- %.3G", np.mean(success),
np.std(success))
pyorg.write_pickle(rands, os.path.join(args.out_path, ver,
"trn_random.pkl"))
del rands[:] # activate garbage collection in loop
def construct_gpn(path, window_sizes):
LOGGER.info("{0:*^78s}".format("Construct GPN"))
version = os.path.basename(path)
if not version:
version = os.path.basename(os.path.dirname(path))
LOGGER.info("{0:*^78s}".format(version))
genes = pyorganism.read_pickle(os.path.join(path, "genes.pkl"))
gpn_gen = pyreg.GPNGenerator()
gpn_gen.parse_genes(genes)
for window in window_sizes:
LOGGER.info("Window Size: %d",window)
gpn = gpn_gen.generate_gpn(window)
LOGGER.info("%d Nodes", len(gpn))
LOGGER.info("%d Links", gpn.size())
components = list(nx.connected_components(gpn))
LOGGER.info("%d Components", len(components))
LOGGER.info("Largest Component: %d", len(components[0]))
if len(components) > 1:
LOGGER.info("Second Largest Component: %d", len(components[1]))
pyorganism.write_pickle(gpn, os.path.join(path, "gpn_%d.pkl" % window))
def main(argv):
LOGGER.info("{0:*^78s}".format("Compile RegulonDB Content"))
if not os.path.exists(argv[0]):
sys.exit(1)
if not os.path.exists(argv[1]):
os.makedirs(argv[1])
version = argv[2] if len(argv) == 3 else os.path.basename(argv[0])
if not version:
version = os.path.basename(os.path.dirname(argv[0]))
LOGGER.info("{0:*^78s}".format(version))
global VERSION
VERSION = version
version = tuple([int(num) for num in version.split(".")])
# compile genes and gene products
gene_file = os.path.join(argv[1], "genes.pkl")
if os.path.exists(gene_file):
genes = pyorganism.read_pickle(gene_file)
else:
genes = compile_genes(argv[0])
product_file = os.path.join(argv[1], "products.pkl")
if os.path.exists(product_file):
products = pyorganism.read_pickle(product_file)
else:
products = compile_products(argv[0])
regdb.link_gene_product(os.path.join(argv[0], "gene_product_link.xml"), VERSION)
tf_file = os.path.join(argv[1], "transcription_factors.pkl")
if os.path.exists(tf_file):
t_factors = pyorganism.read_pickle(tf_file)
else:
t_factors = compile_transcription_factors(argv[0], version)
num = sum(1 for gene in genes if isinstance(gene.regulatory_product, pyreg.TranscriptionFactor))
norm = float(len(genes))
LOGGER.info("Found {0:d} genes that code for transcription factors({1:.2%})".format(num, num / norm))
if version >= (7, 2):
sigma_file = os.path.join(argv[1], "sigma_factors.pkl")
if os.path.exists(sigma_file):
sigma_factors = pyorganism.read_pickle(sigma_file)
else:
sigma_factors = compile_sigma_factors(argv[0])
conf_file = os.path.join(argv[1], "conformations.pkl")
if os.path.exists(conf_file):
conformations = pyorganism.read_pickle(conf_file)
else:
conformations = compile_conformations(argv[0])
prom_file = os.path.join(argv[1], "promoters.pkl")
if os.path.exists(prom_file):
promoters = pyorganism.read_pickle(prom_file)
else:
promoters = compile_promoters(argv[0])
op_file = os.path.join(argv[1], "operons.pkl")
if os.path.exists(op_file):
operons = pyorganism.read_pickle(op_file)
else:
operons = compile_operons(argv[0])
regdb.update_operons(operons, promoters, genes)
tu_file = os.path.join(argv[1], "transcription_units.pkl")
if os.path.exists(tu_file):
t_units = pyorganism.read_pickle(tu_file)
else:
t_units = compile_transcription_units(argv[0])
inter_file = os.path.join(argv[1], "interactions.pkl")
if os.path.exists(inter_file):
interactions = pyorganism.read_pickle(inter_file)
else:
interactions = compile_regulation(argv[0])
# write-out all pickles again since object attributes may have been updated
pyorganism.write_pickle(genes, gene_file)
pyorganism.write_pickle(products, product_file)
pyorganism.write_pickle(t_factors, tf_file)
if version >= (7, 2):
pyorganism.write_pickle(sigma_factors, sigma_file)
pyorganism.write_pickle(conformations, conf_file)
pyorganism.write_pickle(promoters, prom_file)
pyorganism.write_pickle(operons, op_file)
pyorganism.write_pickle(t_units, tu_file)
pyorganism.write_pickle(interactions, inter_file)
