cluster_file = args.cluster_file
cluster_type = args.cluster_type
email = args.email
print_only = args.print_only
verbose = args.verbose
version = "--version " + TopmedPipeline.__version__
cluster = TopmedPipeline.ClusterFactory.createCluster(cluster_type,
cluster_file, verbose)
pipeline = cluster.getPipelinePath()
submitPath = cluster.getSubmitPath()
driver = os.path.join(submitPath, "runRscript.sh")
configdict = TopmedPipeline.readConfig(configfile)
configdict = TopmedPipeline.directorySetup(configdict,
subdirs=["log", "plots"])
# analysis init
cluster.analysisInit(print_only=print_only)
job = "locuszoom"
rscript = os.path.join(pipeline, "R", job + ".R")
# find number of jobs to submit by counting lines in file
n = TopmedPipeline.countLines(configdict["locus_file"])
range = "1-" + str(n - 1)
args = ["-s", rscript, configfile, version]
chromosomes = args.chromosomes
segment_length = args.segment_length
n_segments = args.n_segments
cluster_file = args.cluster_file
cluster_type = args.cluster_type
email = args.email
print_only = args.print_only
verbose = args.verbose
cluster = TopmedPipeline.ClusterFactory.createCluster(cluster_type,
cluster_file, verbose)
pipeline = os.path.dirname(os.path.abspath(sys.argv[0]))
driver = os.path.join(pipeline, "runRscript.sh")
configdict = TopmedPipeline.readConfig(configfile)
configdict = TopmedPipeline.directorySetup(
configdict, subdirs=["config", "data", "log", "plots", "report"])
# check type of association test - single-variant unrelated is handled differently
no_pcrel = "pcrelate_file" not in configdict or configdict[
"pcrelate_file"] == "NA"
no_grm = "grm_file" not in configdict or configdict["grm_file"] == "NA"
single_unrel = assoc_type == "single" and no_pcrel and no_grm
holdids = []
# null model
if not single_unrel:
job = "null_model"
cluster_type = args.cluster_type
ncores = args.ncores
email = args.email
print_only = args.print_only
verbose = args.verbose
version = "--version " + TopmedPipeline.__version__
cluster = TopmedPipeline.ClusterFactory.createCluster(cluster_type,
cluster_file, verbose)
pipeline = cluster.getPipelinePath()
submitPath = cluster.getSubmitPath()
driver = os.path.join(submitPath, "runRscript.sh")
configdict = TopmedPipeline.readConfig(configfile)
configdict = TopmedPipeline.directorySetup(
configdict, subdirs=["config", "data", "log", "report"])
# analysis init
cluster.analysisInit(print_only=print_only)
# null model
job = "null_model"
rscript = os.path.join(pipeline, "R", job + ".R")
config = deepcopy(configdict)
config["out_prefix"] = configdict["data_prefix"] + "_null_model"
config["out_phenotype_file"] = configdict["data_prefix"] + "_phenotypes.RData"
configfile = configdict["config_prefix"] + "_" + job + ".config"
n_segments = args.n_segments
cluster_file = args.cluster_file
cluster_type = args.cluster_type
ncores = args.ncores
email = args.email
print_only = args.print_only
verbose = args.verbose
version = "--version " + TopmedPipeline.__version__
cluster = TopmedPipeline.ClusterFactory.createCluster(cluster_type, cluster_file, verbose)
pipeline = cluster.getPipelinePath()
driver = os.path.join(pipeline, "runRscript.sh")
configdict = TopmedPipeline.readConfig(configfile)
configdict = TopmedPipeline.directorySetup(configdict, subdirs=["config", "data", "log", "plots", "report"])
# hold is a list of submit IDs. A submit ID is a dict:
# {jobname: [jobids]}
hold_null_agg = []
# null model
job = "null_model"
# if a null model file is given in the config, skip this step
run_null_model = "null_model_file" not in configdict
if run_null_model:
rscript = os.path.join(pipeline, "R", job + ".R")
cluster_file = args.cluster_file
cluster_type = args.cluster_type
ncores = args.ncores
email = args.email
print_only = args.print_only
verbose = args.verbose
version = "--version " + TopmedPipeline.__version__
cluster = TopmedPipeline.ClusterFactory.createCluster(cluster_type,
cluster_file, verbose)
pipeline = cluster.getPipelinePath()
driver = os.path.join(pipeline, "runRscript.sh")
configdict = TopmedPipeline.readConfig(configfile)
configdict = TopmedPipeline.directorySetup(
configdict, subdirs=["config", "data", "log", "plots"])
job = "find_unrelated"
rscript = os.path.join(pipeline, "R", job + ".R")
config = deepcopy(configdict)
config["out_related_file"] = configdict["data_prefix"] + "_related.RData"
config["out_unrelated_file"] = configdict["data_prefix"] + "_unrelated.RData"
configfile = configdict["config_prefix"] + "_" + job + ".config"
TopmedPipeline.writeConfig(config, configfile)
jobid = cluster.submitJob(job_name=job,
cmd=driver,
chromosomes = args.chromosomes
#cluster_file = args.cluster_file
#cluster_type = args.cluster_type
ncores = args.ncores
#email = args.email
#print_only = args.print_only
verbose = args.verbose
#version = "--version " + TopmedPipeline.__version__
#cluster = TopmedPipeline.ClusterFactory.createCluster(cluster_type, cluster_file, verbose)
pipeline = os.path.dirname(os.path.abspath(sys.argv[0]))
#driver = os.path.join(pipeline, "runRscript.sh")
configdict = TopmedPipeline.readConfig(configfile)
for subdir in ["config", "data", "log", "plots"]:
if not os.path.exists(configdict['output_file'] + '/' + subdir):
os.mkdir(configdict['output_file'] + '/' + subdir)
#configdict = TopmedPipeline.directorySetup(configdict, subdirs=["config", "data", "log"])
chrom_string = TopmedPipeline.parseChromosomes(chromosomes)
chrom_list = chrom_string.split(' ')
job = "grm"
rscript = os.path.join(pipeline, "R", job + ".R")
config = deepcopy(configdict)
config["out_file"] = configdict['output_file'] + '/data/' + configdict[
"data_prefix"] + "_grm_chr .gds"
cluster_type = args.cluster_type
ncores = args.ncores
email = args.email
print_only = args.print_only
verbose = args.verbose
version = "--version " + TopmedPipeline.__version__
cluster = TopmedPipeline.ClusterFactory.createCluster(cluster_type,
cluster_file, verbose)
pipeline = os.path.dirname(os.path.abspath(sys.argv[0]))
submitPath = cluster.getSubmitPath()
driver = os.path.join(submitPath, "runRscript.sh")
configdict = TopmedPipeline.readConfig(configfile)
configdict = TopmedPipeline.directorySetup(configdict,
subdirs=["config", "log"])
# analysis init
cluster.analysisInit(print_only=print_only)
job = "vcf2gds"
rscript = os.path.join(pipeline, "R", job + ".R")
# parsing bcf files relies on streaming bcftools output, so can't run in parallel
if os.path.splitext(configdict["vcf_file"])[1] == ".bcf":
ncores = None
jobid = cluster.submitJob(job_name=job,
chromosomes = args.chromosomes cluster_file = args.cluster_file cluster_type = args.cluster_type ncores = args.ncores email = args.email print_only = args.print_only verbose = args.verbose version = "--version " + TopmedPipeline.__version__ cluster = TopmedPipeline.ClusterFactory.createCluster(cluster_type, cluster_file, verbose) pipeline = cluster.getPipelinePath() driver = os.path.join(pipeline, "runRscript.sh") configdict = TopmedPipeline.readConfig(configfile) configdict = TopmedPipeline.directorySetup(configdict, subdirs=["config", "data", "log", "plots"]) job = "find_unrelated" rscript = os.path.join(pipeline, "R", job + ".R") config = deepcopy(configdict) config["out_related_file"] = configdict["data_prefix"] + "_related.RData" config["out_unrelated_file"] = configdict["data_prefix"] + "_unrelated.RData" configfile = configdict["config_prefix"] + "_" + job + ".config" TopmedPipeline.writeConfig(config, configfile) jobid = cluster.submitJob(job_name=job, cmd=driver, args=[rscript, configfile, version], email=email, print_only=print_only)
configfile = args.config_file cluster_file = args.cluster_file cluster_type = args.cluster_type email = args.email print_only = args.print_only verbose = args.verbose version = "--version " + TopmedPipeline.__version__ cluster = TopmedPipeline.ClusterFactory.createCluster(cluster_type, cluster_file, verbose) pipeline = cluster.getPipelinePath() driver = os.path.join(pipeline, "runRscript.sh") configdict = TopmedPipeline.readConfig(configfile) configdict = TopmedPipeline.directorySetup(configdict, subdirs=["config", "data", "log", "plots"]) job = "pcrelate" rscript = os.path.join(pipeline, "R", job + ".R") config = deepcopy(configdict) config["out_prefix"] = configdict["data_prefix"] configfile = configdict["config_prefix"] + "_" + job + ".config" TopmedPipeline.writeConfig(config, configfile) jobid = cluster.submitJob(job_name=job, cmd=driver, args=[rscript, configfile, version], email=email, print_only=print_only)
# help="show the version number and exit")
args = parser.parse_args()
assoc_type = args.assoc_type
configfile = args.config_file
chromosomes = args.chromosomes
segment_length = args.segment_length
n_segments = args.n_segments
#email = args.email
#print_only = args.print_only
verbose = args.verbose
#version = "--version " + TopmedPipeline.__version__
pipeline = os.path.dirname(os.path.abspath(sys.argv[0]))
configdict = TopmedPipeline.readConfig(configfile)
for subdir in ["config","data", "results", "log", "plots", "report"]:
if not os.path.exists(configdict['output_file'] + '/' + subdir):
os.mkdir(configdict['output_file'] + '/' + subdir)
# check type of association test - single-variant unrelated is handled differently
no_pcrel = "pcrelate_file" not in configdict or configdict["pcrelate_file"] == "NA"
no_grm = "grm_file" not in configdict or configdict["grm_file"] == "NA"
single_unrel = assoc_type == "single" and no_pcrel and no_grm
#######job: null model#############
if not single_unrel:
job = "null_model"
assocScript = "assoc_" + assoc_type
if "test_type" in configdict:
configfile = args.config_file cluster_file = args.cluster_file cluster_type = args.cluster_type email = args.email print_only = args.print_only verbose = args.verbose version = "--version " + TopmedPipeline.__version__ cluster = TopmedPipeline.ClusterFactory.createCluster(cluster_type, cluster_file, verbose) pipeline = cluster.getPipelinePath() driver = os.path.join(pipeline, "runRscript.sh") configdict = TopmedPipeline.readConfig(configfile) configdict = TopmedPipeline.directorySetup(configdict, subdirs=["log", "plots"]) job = "locuszoom" rscript = os.path.join(pipeline, "R", job + ".R") # find number of jobs to submit by counting lines in file n = TopmedPipeline.countLines(configdict["locus_file"]) range = "1-" + str(n-1) args = ["-s", rscript, configfile, version] jobid = cluster.submitJob(job_name=job, cmd=driver, args=args, array_range=range, email=email, print_only=print_only)
chromosomes = args.chromosomes
cluster_file = args.cluster_file
cluster_type = args.cluster_type
ncores = args.ncores
email = args.email
print_only = args.print_only
verbose = args.verbose
version = "--version " + TopmedPipeline.__version__
cluster = TopmedPipeline.ClusterFactory.createCluster(cluster_type, cluster_file, verbose)
pipeline = os.path.dirname(os.path.abspath(sys.argv[0]))
driver = os.path.join(pipeline, "runRscript.sh")
configdict = TopmedPipeline.readConfig(configfile)
configdict = TopmedPipeline.directorySetup(configdict, subdirs=["config", "log"])
job = "vcf2gds"
rscript = os.path.join(pipeline, "R", job + ".R")
# parsing bcf files relies on streaming bcftools output, so can't run in parallel
if os.path.splitext(configdict["vcf_file"])[1] == ".bcf":
ncores = None
jobid = cluster.submitJob(job_name=job, cmd=driver, args=["-c", rscript, configfile, version], array_range=chromosomes, request_cores=ncores, email=email, print_only=print_only)
job = "check_gds"