def update_mlps(**kwargs):
print('Updating MLPs...')
tasks = []
for lg in models.LocusGroup.objects.all():
for exp_type in models.ExperimentType.objects.all():
if models.Dataset.objects.filter(
assembly=lg.assembly,
experiment__experiment_type=exp_type,
experiment__project__name='ENCODE',
experiment__processed=True,
experiment__revoked=False,
processed=True,
revoked=False,
).count() >= 20:
nn = models.NeuralNetwork.objects.get_or_create(
locus_group=lg,
experiment_type=exp_type,
metadata_field='cell_type',
)[0]
tasks.append(fit_neural_network.si(nn.pk))
nn = models.NeuralNetwork.objects.get_or_create(
locus_group=lg,
experiment_type=exp_type,
metadata_field='target',
)[0]
tasks.append(fit_neural_network.si(nn.pk))
run_tasks(tasks, group_async=True)
print('Done.')
def update_dendrograms(**kwargs):
print('Updating dendrograms...')
tasks = []
for org in models.Organism.objects.all():
for exp_type in models.ExperimentType.objects.all():
experiments = models.Experiment.objects.filter(
organism=org, experiment_type=exp_type)
if experiments.exists():
tasks.append(call_update_dendrogram.si(org.pk, exp_type.pk))
for my_user in models.MyUser.objects.all():
if experiments.filter(owners=my_user).exists():
tasks.append(
call_update_dendrogram.si(org.pk,
exp_type.pk,
my_user_pk=my_user.pk))
run_tasks(tasks, group_async=True)
print('Done.')
def update_networks(**kwargs):
print('Updating network graphs...')
tasks = []
for org in models.Organism.objects.all():
for exp_type in models.ExperimentType.objects.all():
experiments = models.Experiment.objects.filter(
dataset__assembly__organism=org,
experiment_type=exp_type,
)
if experiments.exists():
tasks.append(update_organism_network.si(org.pk, exp_type.pk))
for my_user in models.MyUser.objects.all():
if experiments.filter(owners=my_user).exists():
tasks.append(
update_organism_network.si(org.pk,
exp_type.pk,
my_user_pk=my_user.pk))
run_tasks(tasks, group_async=True)
print('Done.')
def update_similarities(objects, sims_df, update_task, **kwargs):
tasks = []
for obj in objects:
row = sims_df[obj.pk]
other_obj_pks = list([int(val) for val in row.index])
sims = list([bool(val) for val in row])
tasks.append(update_task.si(obj.pk, other_obj_pks, sims))
run_tasks(tasks, **kwargs)
def update_predictions(**kwargs):
print('Updating dataset predictions...')
tasks = []
for ds in models.Dataset.objects.all():
tasks.append(predict_dataset_fields.si(ds.pk))
run_tasks(tasks, group_async=True)
print('Done.')
def update_transcript_selections(**kwargs):
print('Updating transcript selections...')
tasks = []
annotations = models.Annotation.objects.filter(
gene__isnull=False).distinct()
for annotation in annotations:
tasks.append(_set_selected_transcripts_for_genes.si(annotation.pk))
run_tasks(tasks, group_async=True)
print('Done.')
def update_normalized_read_coverage(**kwargs):
print('Updating normalized read coverage...')
tasks = []
dataset_intersections = models.DatasetIntersectionJson.objects.filter(
dataset__processed=True).distinct()
for dij in dataset_intersections:
tasks.append(_normalize_dataset_intersections.si(dij.pk))
run_tasks(tasks, group_async=True)
print('Done.')
def update_feature_values(**kwargs):
print('Updating feature values...')
tasks = []
for dataset in models.Dataset.objects.all():
for locus_group in models.LocusGroup.objects.filter(
assembly=dataset.assembly):
tasks.append(
update_or_create_feature_values.si(dataset.pk, locus_group.pk))
run_tasks(tasks, group_async=True)
print('Done.')
def update_pcas(**kwargs):
print('Updating PCAs...')
tasks = []
for lg in models.LocusGroup.objects.all():
for exp_type in models.ExperimentType.objects.all():
if models.Dataset.objects.filter(
assembly=lg.assembly,
experiment__experiment_type=exp_type,
).count() >= 3: # Verify that there are at least 3 datasets
tasks.append(call_update_pca.si(lg.pk, exp_type.pk))
run_tasks(tasks, group_async=True)
print('Done.')
def update_feature_attributes(**kwargs):
print('Updating feature attributes...')
tasks = []
for locus_group in models.LocusGroup.objects.all():
for experiment_type in models.ExperimentType.objects.all():
if models.Dataset.objects.filter(
assembly=locus_group.assembly,
experiment__experiment_type=experiment_type,
experiment__project__isnull=False,
).exists():
tasks.append(
update_or_create_feature_attributes.si(
locus_group.pk, experiment_type.pk))
run_tasks(tasks, group_async=True)
print('Done.')
def update_experiment_list_recommendations(experiments, **kwargs):
tasks = []
for exp in experiments:
tasks.append(update_experiment_recommendations.si(exp.pk, **kwargs))
run_tasks(tasks, **kwargs)