def run_permutation(self, seeds):
"""
performs permutations
"""
permuted_pheno_file = common.create_pheno_permutation_file(self, seeds)
pheno_arg = " --pheno " + str(permuted_pheno_file + " --all-pheno" + " --extract " + self.group)
if self.covar_file:
permuted_covar_file = common.create_covar_permutation_file(self, seeds)
pheno_arg += " --covar " + permuted_covar_file + " --hide-covar"
self.set_plink_arguments("--bfile " + self.bfile + pheno_arg)
else:
self.set_plink_arguments("--bfile " + self.bfile + " --assoc " + pheno_arg)
resultfile = self.run_plink() # run plink
results_extension = common.plink_results_extension(resultfile)
resultfiles = self.point_outputfile_to_phenotype(seeds, resultfile, results_extension)
results = extract_permutated_scores(self.snp_to_group_map, resultfiles, seeds)
os.remove(permuted_pheno_file) #remove permuted pheno file
common.remove_plink_output(resultfile) # remove all output from plink
return results
def run_step1(self, plink, geneset, gene_group=False, adjusted=False):
"""
main function of association analysis. this function calls all function
needed for aa analysis.
plink. a plink object. Example to create this kind of object can be found in test/test_jag.py
group: path to file which contains file mapping
"""
#check files are available
#read geneset as dictionary (key=snp, value=geneset_name)
if (not geneset == None):
snptogeneset = common.map_snp_to_geneset(geneset, gene_group)
pheno_arg = ""
if (not plink.pheno_file == None):
pheno_arg = " --pheno " + str(plink.pheno_file + " --all-pheno ")
if not plink.covar_file == None:
pheno_arg += " --covar " + plink.covar_file + " --hide-covar "
plink.set_plink_arguments("--bfile " + plink.bfile + " " + pheno_arg)
else:
plink.set_plink_arguments("--bfile " + plink.bfile + " --assoc " + pheno_arg)
resultfile = plink.run_plink()
assoc_files = glob.glob(resultfile + "*.*assoc*")
if adjusted:
assoc_files = [assoc_file for assoc_file in assoc_files if assoc_file.endswith("adjusted")]
else:
assoc_files = [assoc_file for assoc_file in assoc_files if not assoc_file.endswith("adjusted")]
pheno_nr=1
for assoc_file in assoc_files:
clusterresults = CR.Clusterresults() # read values from assoc assoc_file and lookup in snptogroup map:
clusterresults.pheno_path = plink.pheno_file
clusterresults.total_columns = plink.get_number_of_pheno_types()
clusterresults.column_number = str(in_out.get_assoc_file_number(resultfile, assoc_file))
self.files[clusterresults.column_number] = {}
clusterresults.bfile_path = plink.bfile
clusterresults.geneset_path = geneset
moved_assoc_file = self.in_and_out.copy_assoc_file(assoc_file)
if (not geneset == None):
clusterresults.map_p_values_from_assoc_file(snptogeneset, assoc_file, adjusted) # extract need p values from assoc file
self.files[clusterresults.column_number]["raw_p_files"] = clusterresults.write_all_pvalues()
clusterresults.createcleanresults()
self.files[clusterresults.column_number]["clusterresults"] = clusterresults
self.in_and_out.save_sumlog(assoc_file, clusterresults) #'generate .sumlog'
else:
log.info ("No geneset file is set: no sumlog file is made.")
sys.exit("JAG is terminated.")
# code for QQ plots
self.files[clusterresults.column_number]["assoc_files"] = moved_assoc_file #saving full file path for creating plot later on
assoc_pheno = [values["assoc_files"] for values in self.files.itervalues()]
assoc_files = [ assoc_file for assoc_file in plot_with_r.from_iterable(assoc_pheno)]
raw_p_pheno = [values["raw_p_files"] for values in self.files.itervalues()]
set_files = [ assoc_file for assoc_file in plot_with_r.from_iterable(raw_p_pheno)]
if self.in_and_out.run_rproject:
plotter = plot_with_r.call_r(self.in_and_out)
plotter.draw_qq_plots(assoc_files, pheno_nr, adjusted)
plotter.draw_qq_plots_for_sets(raw_p_pheno, pheno_nr)
pheno_nr=pheno_nr+1
common.remove_plink_output(resultfile) # remove all output from plink
return self.files