def localpermutation(self, nperm, random_seed, gene_group=False):
'''
Run permutation on local computer
'''
self.gene_based = gene_group
aa_results = self.get_assoc_analysis()
#create a random seed
if (random_seed == None):
random_seed = _create_random_seed(8)
snp_to_group = common.map_snp_to_geneset(self.group, gene_group)
self.plink.set_snp_to_group_map(snp_to_group)
nperm = int(nperm)
permperjob = int(50)
if self.plink.covar_file:
permperjob = 1
jobs = _create_seeds_for_jobs(nperm, permperjob, random_seed)
resultfiles = self.plink.run_permutation_single_core(jobs)
results_merged = merge_results(resultfiles)
for key, perm_score_per_group in results_merged.iteritems():
perm_out_string = perm_score_per_group.format_permout()
perm_out_filename = str(random_seed) + ".P" + str(key) + ".perm"
#save permutation score as table with random seed in filename
perm_filename = self.in_out.save_text_to_filename(perm_out_filename, perm_out_string)
log.info("\nSaved permutation results as " + perm_filename)
key = str(key)
self.files[key] = {"perm":perm_filename}
if (self.no_emp == False):
column_aaresults_mapping = {}
for i in range(len(aa_results)):
column_aaresults_mapping[str(aa_results[i].column_number)] = i
current_aa_results = aa_results[column_aaresults_mapping[key]]
empp_out_text = perm_score_per_group.format_permutated_results(current_aa_results)
emp_filename = "P" + key + ".empp"
emp_filename = self.in_out.save_text_to_filename(emp_filename, empp_out_text)
log.info("Saved empirical pvalues as " + emp_filename)
self.files[str(key)] ["empp"] = emp_filename
if self.in_out.run_rproject:
import jag.plot_with_r as plot_with_r
plotter = plot_with_r.call_r(self.in_out)
plotter.draw_dist_plot(self.files, key)
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
