def shared_to_directory(population, labeled_nodes, genome_generator,
recombinators, directory, min_segment_length = 0,
clobber = True, iterations = 1000,
generations_back_shared = 7):
labeled_nodes = set(labeled_nodes)
unlabeled_nodes = chain.from_iterable(generation.members
for generation
in population.generations[-3:])
unlabeled_nodes = set(unlabeled_nodes) - labeled_nodes
print("Finding related pairs.")
pairs = related_pairs(unlabeled_nodes, labeled_nodes, population,
generations_back_shared)
print("{} related pairs.".format(len(pairs)))
print("Opening file descriptors.")
if clobber:
mode = "w"
else:
mode = "a"
fds = {node: open(join(directory, str(node._id)), mode)
for node in labeled_nodes}
print("Calculating shared lengths.")
for i in range(iterations):
print("iteration {}".format(i))
print("Cleaning genomes.")
population.clean_genomes()
print("Generating genomes")
generate_genomes(population, genome_generator, recombinators, 3)
print("Calculating shared length")
_calculate_shared_to_fds(pairs, fds, min_segment_length)
for fd in fds.values():
fd.close()
def simulate_founder_stats(population, genome_generator, recombinators,
iterations, output_file):
node_stats = defaultdict(lambda: defaultdict(list))
for i in range(iterations):
genome_generator.reset()
print("iteration {}".format(i))
print("Cleaning genomes.")
population.clean_genomes()
print("Generating genomes")
generate_genomes(population,
genome_generator,
recombinators,
3,
true_genealogy=False)
print("Calculating founder stats")
nodes = (node for node in population.members
if node.genome is not None)
for node in nodes:
mother = node.genome.mother
father = node.genome.father
temp_founder_lengths = defaultdict(list)
diploid_founders(mother, temp_founder_lengths)
diploid_founders(father, temp_founder_lengths)
founder_map = node_stats[str(node._id)]
for founder, lengths in temp_founder_lengths.items():
founder_map[str(founder)].append((sum(lengths), len(lengths)))
with open(output_file, "w") as output_json:
dump(node_stats, output_json)
def shared_to_directory(population,
labeled_nodes,
genome_generator,
recombinators,
directory,
min_segment_length=0,
clobber=True,
iterations=1000,
generations_back_shared=7,
non_paternity=0.0):
labeled_nodes = set(labeled_nodes)
unlabeled_nodes = chain.from_iterable(
generation.members for generation in population.generations[-3:])
unlabeled_nodes = set(unlabeled_nodes) # - labeled_nodes
print("Finding related pairs.")
pairs = related_pairs(unlabeled_nodes, labeled_nodes, population,
generations_back_shared)
print("{} related pairs.".format(len(pairs)))
print("Opening file descriptors.")
if clobber:
mode = "w"
else:
mode = "a"
fds = {
node: open(join(directory, str(node._id)), mode)
for node in labeled_nodes
}
suppressor = ParentSuppressor(non_paternity, 0.0)
print("Calculating shared lengths.")
for i in range(iterations):
print("iteration {}".format(i))
print("Cleaning genomes.")
population.clean_genomes()
# print("Perturbing parentage")
# suppressor.suppress(population)
print("Generating genomes")
generate_genomes(population,
genome_generator,
recombinators,
3,
true_genealogy=False)
print("Calculating shared length")
_calculate_shared_to_fds(pairs, fds, min_segment_length)
# print("Fixing perturbation")
# suppressor.unsuppress()
for fd in fds.values():
fd.close()
def shared_to_directory(population,
labeled_nodes,
genome_generator,
recombinators,
directory,
min_segment_length=0,
clobber=True,
iterations=1000,
generations_back_shared=7):
labeled_nodes = set(labeled_nodes)
unlabeled_nodes = chain.from_iterable(
generation.members for generation in population.generations[-3:])
unlabeled_nodes = set(unlabeled_nodes) - labeled_nodes
print("Finding related pairs.")
pairs = related_pairs(unlabeled_nodes, labeled_nodes, population,
generations_back_shared)
print("{} related pairs.".format(len(pairs)))
print("Opening file descriptors.")
if clobber:
mode = "w"
else:
mode = "a"
fds = {
node: open(join(directory, str(node._id)), mode)
for node in labeled_nodes
}
print("Calculating shared lengths.")
for i in range(iterations):
print("iteration {}".format(i))
print("Cleaning genomes.")
population.clean_genomes()
print("Generating genomes")
generate_genomes(population, genome_generator, recombinators, 3)
print("Calculating shared length")
_calculate_shared_to_fds(pairs, fds, min_segment_length)
for fd in fds.values():
fd.close()
tree = tree_from_file(args.tree_file)
leaves = tree.leaves
for person in founders:
tree.add_individual(choice(leaves), person)
population = HierarchicalIslandPopulation(tree)
for _ in range(args.num_generations - 1):
population.new_generation()
if not args.no_genomes:
# tr = tracker.SummaryTracker()
recombinators = recombinators_from_directory(args.recombination_dir)
chrom_sizes = recombinators[Sex.Male]._num_bases
genome_generator = RecombGenomeGenerator(chrom_sizes)
generate_genomes(population, genome_generator, recombinators, 3)
# tr.print_diff()
# summary.print_(summary.summarize(muppy.get_objects()))
# genomes = [m.genome for m in population.members]
# print("genome sizes: " + str(asizeof.asizeof(genomes) // 1024))
# print("population size: " + str(asizeof.asizeof(population) // 1024))
if args.output_file:
with open(args.output_file, "wb") as pickle_file:
# Trees cause deep recursion in the pickle module, so we need
# to raise the recursion limit. This is the stack depth for
# python functions, you may need to increase the native stack
# depth using ulimit -s
# https://docs.python.org/3.4/library/pickle.html#what-can-be-pickled-and-unpickled
dump(population, pickle_file, protocol = HIGHEST_PROTOCOL)
