def choose_best(self, phylogenetics):
# Get all the phylo_groups present.
phylo_groups = self._get_present_phylo_groups(
phylogenetics["phylo_group"])
phylogenetics["phylo_group"] = phylo_groups
sub_complexes = self._get_present_phylo_groups(
phylogenetics["sub_complex"], mix_threshold=90)
phylogenetics["sub_complex"] = sub_complexes
# for each phylo_group, get the best species in the phylo_group (using
# sub_complex info where possible)
species = {}
for pg in phylo_groups.keys():
if self.hierarchy:
allowed_species = flatten([
self.hierarchy.dict[pg]["children"][subc]
["children"].keys()
for subc in self.hierarchy.dict[pg]["children"].keys()
if subc is not "Unknown"
])
species_to_consider = {
k: phylogenetics["species"].get(k, {"percent_coverage": 0})
for k in allowed_species
}
else:
species_to_consider = phylogenetics["species"]
best_species = self._get_present_phylo_groups(species_to_consider,
mix_threshold=90)
species.update(best_species)
phylogenetics["species"] = species
# For each species, get the best sub species where applicable
sub_species = {}
for s in species.keys():
if self.hierarchy:
allowed_sub_species = self.hierarchy.get_children(s)
sub_species_to_consider = {
k: phylogenetics["lineage"].get(k, {"percent_coverage": 0})
for k in allowed_sub_species
}
else:
sub_species_to_consider = phylogenetics.get("lineage", {})
best_sub_species = self._get_best_coverage_dict(
sub_species_to_consider)
sub_species.update(best_sub_species)
phylogenetics["lineage"] = sub_species
return phylogenetics
def seen_together(variants):
# Takes a list of variants.
# Returns a list of variants that appear together (in the same variant set)
variant_id_to_samples = {}
for variant in variants:
variant_id_to_samples[variant.var_hash] = variant.seen_in_samples()
samples_counter = Counter(flatten(variant_id_to_samples.values()))
samples_seen_more_than_once = [
k for k, v in samples_counter.items() if v > 1
]
contexts = []
for sample in samples_seen_more_than_once:
vars_together = []
for variant_id, samples in variant_id_to_samples.items():
if sample in samples:
vars_together.append(variant_id)
if vars_together not in contexts:
contexts.append(vars_together)
variants = [
var for var in variants if var.var_hash not in vars_together
]
for var in variants:
contexts.append([var.var_hash])
new_contexts = []
for context in contexts:
new_context = []
for variant_id in context:
new_context.append(Variant.objects.get(var_hash=variant_id))
new_contexts.append(new_context)
return new_contexts + [[]]
def _get_drug_list_from_variant_to_resistance_drug(self):
return unique(
flatten(self.variant_or_gene_name_to_resistance_drug.values()))
def _get_drug_list_from_variant_to_resistance_drug(self):
return unique(flatten(self.variant_or_gene_name_to_resistance_drug.values()))