MPControl.set_multiprocess_engine("dask-cluster")
MPControl.client.use_default_configuration("rusty_ccb", n_jobs=n_jobs)
MPControl.client.add_worker_conda(
"source ~/.local/anaconda3/bin/activate inferelator")
MPControl.client.add_slurm_command_line("--constraint=broadwell")
MPControl.connect()
if __name__ == '__main__':
set_up_dask()
utils.Debug.vprint("Testing preprocessing", level=0)
# Figure 5D: BBSR By Task Learning
worker = workflow.inferelator_workflow(regression="amusr",
workflow="multitask")
set_up_workflow(worker)
worker.append_to_path('output_dir', 'figure_4_count')
cv_wrap = set_up_cv_seeds(worker)
cv_wrap.run()
del cv_wrap
del worker
worker = workflow.inferelator_workflow(regression="amusr",
workflow="multitask")
set_up_workflow(worker)
worker.add_preprocess_step("log2")
worker.append_to_path('output_dir', 'figure_4_log2')
cv_wrap = set_up_cv_seeds(worker)
cv_wrap.run()
MPControl.set_multiprocess_engine("dask-cluster")
MPControl.client.use_default_configuration("rusty_ccb", n_jobs=n_jobs)
MPControl.client.add_worker_conda(
"source ~/.local/anaconda3/bin/activate inferelator")
MPControl.client.add_slurm_command_line("--constraint=broadwell")
MPControl.connect()
if __name__ == '__main__':
set_up_dask()
utils.Debug.vprint("Testing preprocessing", level=0)
# Figure 5D: BBSR By Task Learning
worker = workflow.inferelator_workflow(regression="bbsr-by-task",
workflow="multitask")
set_up_workflow(worker)
worker.append_to_path('output_dir', 'figure_4_count')
cv_wrap = set_up_cv_seeds(worker)
#cv_wrap.run()
del cv_wrap
del worker
worker = workflow.inferelator_workflow(regression="bbsr-by-task",
workflow="multitask")
set_up_workflow(worker)
worker.add_preprocess_step("log2")
worker.append_to_path('output_dir', 'figure_4_log2')
cv_wrap = set_up_cv_seeds(worker)
#cv_wrap.run()
priors_file=PRIORS_FILE_NAME,
gold_standard_file=GOLD_STANDARD_FILE_NAME)
wkf.set_expression_file(tsv=EXPRESSION_FILE_NAME)
wkf.set_file_properties(expression_matrix_columns_are_genes=False)
wkf.set_run_parameters(num_bootstraps=5)
wkf.set_crossvalidation_parameters(split_gold_standard_for_crossvalidation=True, cv_split_ratio=0.2)
return wkf
# Inference with BBSR (crossvalidation)
# Using the crossvalidation wrapper
# Run the regression 10 times and hold 20% of the gold standard out of the priors for testing each time
# Each run is seeded differently (and therefore has different holdouts)
# Create a worker
worker = workflow.inferelator_workflow(regression="bbsr", workflow="tfa")
worker = set_up_workflow(worker)
worker.append_to_path("output_dir", "bbsr")
# Create a crossvalidation wrapper
cv_wrap = CrossValidationManager(worker)
# Assign variables for grid search
cv_wrap.add_gridsearch_parameter('random_seed', CV_SEEDS)
# Run
cv_wrap.run()
# Inference with Elastic Net (crossvalidation)
# Using the crossvalidation wrapper
# Run the regression 10 times and hold 20% of the gold standard out of the priors for testing each time
MPControl.set_multiprocess_engine("dask-cluster")
MPControl.client.use_default_configuration("rusty_ccb", n_jobs=n_jobs)
MPControl.client.add_worker_conda(
"source ~/.local/anaconda3/bin/activate inferelator")
MPControl.client.add_slurm_command_line("--constraint=broadwell")
MPControl.connect()
if __name__ == '__main__':
set_up_dask()
utils.Debug.vprint("Testing preprocessing", level=0)
# Figure 5D: BBSR By Task Learning
worker = workflow.inferelator_workflow(regression="bbsr",
workflow="single-cell")
set_up_workflow(worker)
worker.append_to_path('output_dir', 'figure_4_count')
cv_wrap = set_up_cv_seeds(worker)
cv_wrap.run()
del cv_wrap
del worker
worker = workflow.inferelator_workflow(regression="bbsr",
workflow="single-cell")
set_up_workflow(worker)
worker.add_preprocess_step("log2")
worker.append_to_path('output_dir', 'figure_4_log2')
cv_wrap = set_up_cv_seeds(worker)
cv_wrap.run()
YEASTRACT_TF_NAMES = "tf_names_yeastract.txt"
INPUT_DIR = '/mnt/ceph/users/cjackson/inferelator/data/yeast'
OUTPUT_PATH = '/mnt/ceph/users/cjackson/jackson_2019_inferelator_v050/'
OUTPUT_FOLDER = "tau43_c"
if __name__ == '__main__':
MPControl.set_multiprocess_engine("dask-cluster")
MPControl.client.use_default_configuration("rusty_preempt")
MPControl.client.set_job_size_params(n_jobs=1)
MPControl.client.add_worker_conda(
"source ~/.local/anaconda3/bin/activate inferelator")
MPControl.connect()
wkf = inferelator_workflow("stars", VelocityWorkflow)
wkf.set_file_paths(input_dir=INPUT_DIR,
output_dir=os.path.join(OUTPUT_PATH, OUTPUT_FOLDER),
gold_standard_file='gold_standard.tsv',
priors_file=YEASTRACT_PRIOR,
tf_names_file=YEASTRACT_TF_NAMES)
wkf.set_expression_file(h5ad=DATA_FILE, h5_layer="counts")
wkf.set_velocity_parameters(velocity_file_name=DATA_FILE,
velocity_file_type="h5ad",
velocity_file_layer="pv")
wkf.set_crossvalidation_parameters(
split_gold_standard_for_crossvalidation=True, cv_split_ratio=0.5)
wkf.set_run_parameters(num_bootstraps=5)
wkf.set_count_minimum(0.05)
wkf.add_preprocess_step(single_cell.log2_data)
wkf.tau = 43.28
def setUp(self):
self.worker = inferelator_workflow()
from inferelator import workflow
from inferelator.distributed.inferelator_mp import MPControl
from inferelator import utils
utils.Debug.set_verbose_level(1)
MPControl.set_multiprocess_engine("multiprocessing")
MPControl.client.processes = 3
wflow = workflow.inferelator_workflow(regression="bbsr", workflow="tfa")
# Common configuration parameters
wflow.input_dir = 'data/bsubtilis'
wflow.num_bootstraps = 2
wflow.delTmax = 110
wflow.delTmin = 0
wflow.tau = 45
if __name__ == "__main__":
wflow.run()
MPControl.set_multiprocess_engine("dask-cluster")
MPControl.client.minimum_cores = n_cores
MPControl.client.maximum_cores = n_cores
MPControl.client.walltime = '48:00:00'
MPControl.client.add_worker_env_line('module load slurm')
MPControl.client.add_worker_env_line('module load gcc/8.3.0')
MPControl.client.add_worker_env_line('source ' + CONDA_ACTIVATE_PATH)
MPControl.client.cluster_controller_options.append("-p ccb")
MPControl.connect()
if __name__ == '__main__':
start_mpcontrol_dask(100)
for seed in range(42, 52):
worker = workflow.inferelator_workflow(regression="amusr",
workflow="amusr")
worker.set_file_paths(input_dir=INPUT_DIR,
output_dir=OUTPUT_DIR,
gold_standard_file="gold_standard.tsv",
gene_metadata_file="orfs.tsv",
priors_file=YEASTRACT_PRIOR,
tf_names_file=YEASTRACT_TF_NAMES)
worker.set_file_properties(gene_list_index="SystematicName")
worker.set_task_filters(target_expression_filter="union",
regulator_expression_filter="intersection")
worker.set_run_parameters(num_bootstraps=5, random_seed=seed)
worker.set_crossvalidation_parameters(
split_gold_standard_for_crossvalidation=True, cv_split_ratio=0.2)
worker.append_to_path('output_dir',
def test_tfa(self):
from inferelator.tfa_workflow import TFAWorkFlow
worker = workflow.inferelator_workflow(regression=RegressionWorkflow,
workflow="tfa")
self.assertTrue(isinstance(worker, TFAWorkFlow))
def test_singlecell(self):
from inferelator.single_cell_workflow import SingleCellWorkflow
worker = workflow.inferelator_workflow(regression=RegressionWorkflow,
workflow="single-cell")
self.assertTrue(isinstance(worker, SingleCellWorkflow))
def test_elasticnet(self):
from inferelator.regression.elasticnet_python import ElasticNetWorkflow
worker = workflow.inferelator_workflow(regression="elasticnet",
workflow=workflow.WorkflowBase)
self.assertTrue(isinstance(worker, ElasticNetWorkflow))
def test_bbsr(self):
from inferelator.regression.bbsr_python import BBSRRegressionWorkflow
worker = workflow.inferelator_workflow(regression="bbsr",
workflow=workflow.WorkflowBase)
self.assertTrue(isinstance(worker, BBSRRegressionWorkflow))
def test_base(self):
worker = workflow.inferelator_workflow(regression=None,
workflow=workflow.WorkflowBase)
with self.assertRaises(NotImplementedError):
worker.run()
TF_NAMES = "tf_names_gold_standard.txt"
YEASTRACT_TF_NAMES = "tf_names_yeastract.txt"
INPUT_DIR = '/mnt/ceph/users/cjackson/inferelator/data/yeast'
OUTPUT_PATH = '/mnt/ceph/users/cjackson/jackson_2019_inferelator_v050/'
if __name__ == '__main__':
MPControl.set_multiprocess_engine("dask-cluster")
MPControl.client.use_default_configuration("rusty_preempt")
MPControl.client.set_job_size_params(n_jobs=1)
MPControl.client.add_worker_conda(
"source ~/.local/anaconda3/bin/activate inferelator")
MPControl.connect()
wkf = inferelator_workflow("stars", "single-cell")
wkf.set_file_paths(input_dir=INPUT_DIR,
output_dir=os.path.join(OUTPUT_PATH, "no_velocity"),
gold_standard_file='gold_standard.tsv',
priors_file=YEASTRACT_PRIOR,
tf_names_file=YEASTRACT_TF_NAMES)
wkf.set_expression_file(h5ad=DATA_FILE, h5_layer="smooth_count")
wkf.set_crossvalidation_parameters(
split_gold_standard_for_crossvalidation=True, cv_split_ratio=0.5)
wkf.set_run_parameters(num_bootstraps=5)
wkf.set_count_minimum(0.05)
wkf.add_preprocess_step(single_cell.log2_data)
cv_wrap = crossvalidation_workflow.CrossValidationManager(wkf)
cv_wrap.add_gridsearch_parameter('random_seed', list(range(42, 52)))
metadata_handler="nonbranching")
wkf.set_run_parameters(num_bootstraps=5)
wkf.set_crossvalidation_parameters(
split_gold_standard_for_crossvalidation=True, cv_split_ratio=0.2)
wkf.append_to_path('output_dir', "yeast_calico")
if __name__ == '__main__':
MPControl.set_multiprocess_engine("dask-cluster")
MPControl.client.use_default_configuration("rusty_ccb", n_jobs=3)
MPControl.client.add_worker_conda(
"source ~/.local/anaconda3/bin/activate inferelator")
MPControl.connect()
worker = workflow.inferelator_workflow(regression="stars",
workflow="single-cell")
single_cell_setup(worker)
# Create a crossvalidation wrapper
cv_wrap = CrossValidationManager(worker)
# Assign variables for grid search
cv_wrap.add_gridsearch_parameter('random_seed', list(range(42, 52)))
# Run
# cv_wrap.run()
del worker
del cv_wrap
worker = workflow.inferelator_workflow(regression="stars", workflow="tfa")
calico_setup(worker)
MPControl.set_multiprocess_engine("dask-cluster")
MPControl.client.minimum_cores = n_cores
MPControl.client.maximum_cores = n_cores
MPControl.client.walltime = '48:00:00'
MPControl.client.add_worker_env_line('module load slurm')
MPControl.client.add_worker_env_line('module load gcc/8.3.0')
MPControl.client.add_worker_env_line('source ' + CONDA_ACTIVATE_PATH)
MPControl.client.cluster_controller_options.append("-p ccb")
MPControl.connect()
if __name__ == '__main__':
start_mpcontrol_dask(100)
for seed in range(42, 52):
worker = workflow.inferelator_workflow(
regression=BBSRByTaskRegressionWorkflow, workflow="amusr")
worker.set_file_paths(input_dir=INPUT_DIR,
output_dir=OUTPUT_DIR,
gold_standard_file="gold_standard.tsv",
gene_metadata_file="orfs.tsv",
priors_file=YEASTRACT_PRIOR,
tf_names_file=YEASTRACT_TF_NAMES)
worker.set_file_properties(gene_list_index="SystematicName")
worker.set_task_filters(target_expression_filter="union",
regulator_expression_filter="intersection")
worker.set_run_parameters(num_bootstraps=5, random_seed=seed)
worker.set_crossvalidation_parameters(
split_gold_standard_for_crossvalidation=True, cv_split_ratio=0.2)
worker.append_to_path('output_dir',
"hybrid_bbsr_mtl_seed_" + str(seed))
def setUp(self):
self.workflow = workflow.inferelator_workflow(workflow="amusr",
regression="amusr")
self.workflow.create_output_dir = lambda *x: None
self.workflow.gold_standard = TaskDataStub.priors_data.copy()
