def buildBetas(self, weight_db_logic, name):
output_path = os.path.join(self.output_folder, name)
if not ".gz" in output_path:
output_path += ".gz"
if os.path.exists(output_path):
logging.info(
"%s already exists, delete it if you want it to be done again",
output_path)
return
logging.info("Building beta for %s and %s", name, self.weight_db_path)
input_path = os.path.join(self.gwas_folder, name)
file_format = GWASUtilities.GWASFileFormat.fileFormatFromArgs(
input_path, self.args)
scheme = MethodGuessing.chooseGWASProcessingScheme(
self.args, input_path)
callback = MethodGuessing.chooseGWASCallback(file_format, scheme,
weight_db_logic)
if not weight_db_logic:
GWASUtilities.loadGWASAndStream(input_path, output_path,
self.compressed_gwas,
self.args.separator,
self.args.skip_until_header,
callback)
else:
dosage_loader = GWASUtilities.GWASDosageFileLoader(
input_path, self.compressed_gwas, self.args.separator,
self.args.skip_until_header, callback)
results, column_order = dosage_loader.load()
# The following check is sort of redundant, as it exists in "saveSetsToCompressedFile".
# It exists merely to provide different logging
if len(results):
def do_output(file, results, column_order):
file.write("\t".join(column_order) + "\n")
first = results[column_order[0]]
n = len(first)
for i in xrange(0, n):
line_comps = [str(results[c][i]) for c in column_order]
line = "%s\n" % "\t".join(line_comps)
file.write(line)
with gzip.open(output_path, "wb") as file:
do_output(file, results, column_order)
else:
logging.info(
"No snps from the tissue model found in the GWAS file")
logging.info("Successfully ran GWAS input processing")
def run(args):
logging.info("Loading weight db")
weight_db_logic = WeightDBUtilities.WeightDBEntryLogic(args.weight_db_path)
logging.info("Loading covariance file")
covariance_contents = MatrixUtilities.loadMatrixFromFile(args.covariance)
logging.info("Choosing method")
beta_contents = Utilities.contentsWithPatternsFromFolder(args.beta_folder, [])
zscore_calculation, normalization = MethodGuessing.chooseZscoreSchemeFromFiles(args.beta_folder, beta_contents, covariance_contents, weight_db_logic)
logging.info("Processing")
betas = {}
for content in beta_contents:
logging.info("Loading betas")
beta_path = os.path.join(args.beta_folder, content)
beta_sets = KeyedDataSet.KeyedDataSetFileUtilities.loadDataSetsFromCompressedFile(beta_path, header="")
beta_sets = {set.name: set for set in beta_sets}
betas[content] = beta_sets
if args.gene_name:
try:
gene_data, weights, covariance_matrix, valid_rsids, beta_sets = get_gene_data(args.gene_name, weight_db_logic, covariance_contents, betas)
weight_values, variances = ZScoreCalculation.preProcess(covariance_matrix, valid_rsids, weights, beta_sets)
if args.interactive:
embed()
logging.info("Processed gene data")
except Exception as e:
logging.info("Couldn't get gene data")
embed()
def buildBetas(self, weight_db_logic, name):
output_path = os.path.join(self.output_folder, name)
if os.path.exists(output_path):
logging.info(
"%s already exists, delete it if you want it to be done again",
output_path)
return
logging.info("Building beta for %s and %s", name, self.weight_db_path)
input_path = os.path.join(self.gwas_folder, name)
file_format = GWASUtilities.GWASFileFormat.fileFormatFromArgs(
input_path, self.args)
scheme = MethodGuessing.chooseGWASProcessingScheme(
file_format, weight_db_logic, self.args, input_path)
callback = GWASUtilities.GWASWeightDBFilteredBetaLineCollector(
file_format, scheme, weight_db_logic)
dosage_loader = GWASUtilities.GWASDosageFileLoader(
input_path, self.compressed, self.args.separator,
self.args.skip_until_header, callback)
result_sets = dosage_loader.load()
# The following check is sort of redundant, as it exists in "saveSetsToCompressedFile".
# It exists merely to provide different login
if len(result_sets):
KeyedDataSet.KeyedDataSetFileUtilities.saveSetsToCompressedFile(
output_path, result_sets, "rsid")
else:
logging.info(
"No snps from the tissue model found in the GWAS file")
def selectMethod(self, folder, beta_contents, covariance_entries, weight_db_logic):
normalization = None
zscore_calculation = None
if self.zscore_scheme:
zscore_calculation = ZScoreCalculation.ZScoreScheme(self.zscore_scheme)
if not self.normalization_scheme:
raise Exception("Normalization scheme is required")
else:
zscore_calculation, normalization = MethodGuessing.chooseZscoreSchemeFromFiles(folder, beta_contents, covariance_entries, weight_db_logic)
if self.normalization_scheme:
normalization = Normalization.normalizationScheme(self.normalization_scheme, covariance_entries, weight_db_logic)
return zscore_calculation, normalization
def buildBetas(self, weight_db_logic, name):
output_path = os.path.join(self.output_folder, name)
if not ".gz" in output_path:
output_path += ".gz"
if os.path.exists(output_path):
logging.info("%s already exists, delete it if you want it to be done again", output_path)
return
logging.info("Building beta for %s and %s", name, self.weight_db_path if self.weight_db_path else "no database")
input_path = os.path.join(self.gwas_folder, name)
file_format = GWASUtilities.GWASFileFormat.fileFormatFromArgs(input_path, self.args)
scheme = MethodGuessing.chooseGWASProcessingScheme(self.args, input_path)
callback = MethodGuessing.chooseGWASCallback(file_format, scheme, weight_db_logic)
if not weight_db_logic:
GWASUtilities.loadGWASAndStream(input_path, output_path, self.compressed_gwas, self.args.separator, self.args.skip_until_header, callback)
else:
dosage_loader = GWASUtilities.GWASDosageFileLoader(input_path, self.compressed_gwas, self.args.separator, self.args.skip_until_header, callback)
results, column_order = dosage_loader.load()
# The following check is sort of redundant, as it exists in "saveSetsToCompressedFile".
# It exists merely to provide different logging
if len(results):
def do_output(file, results, column_order):
file.write("\t".join(column_order)+"\n")
first = results[column_order[0]]
n = len(first)
for i in xrange(0,n):
line_comps = [str(results[c][i]) for c in column_order]
line = "%s\n" % "\t".join(line_comps)
file.write(line)
with gzip.open(output_path, "wb") as file:
do_output(file, results, column_order)
else:
logging.info("No snps from the tissue model found in the GWAS file")
logging.info("Successfully ran GWAS input processing")
def selectMethod(self, folder, beta_contents, covariance_entries,
weight_db_logic):
normalization = None
zscore_calculation = None
if self.zscore_scheme:
zscore_calculation = ZScoreCalculation.ZScoreScheme(
self.zscore_scheme)
if not self.normalization_scheme:
raise Exception("Normalization scheme is required")
else:
zscore_calculation, normalization = MethodGuessing.chooseZscoreSchemeFromFiles(
folder, beta_contents, covariance_entries, weight_db_logic)
if self.normalization_scheme:
normalization = Normalization.normalizationScheme(
self.normalization_scheme, covariance_entries, weight_db_logic)
return zscore_calculation, normalization
def buildBetas(self, weight_db_logic, name):
output_path = os.path.join(self.output_folder, name)
if os.path.exists(output_path):
logging.info("%s already exists, delete it if you want it to be done again", output_path)
return
logging.info("Building beta for %s and %s", name, self.weight_db_path)
input_path = os.path.join(self.gwas_folder, name)
file_format = GWASUtilities.GWASFileFormat.fileFormatFromArgs(input_path, self.args)
scheme = MethodGuessing.chooseGWASProcessingScheme(file_format, weight_db_logic, self.args, input_path)
callback = GWASUtilities.GWASWeightDBFilteredBetaLineCollector(file_format, scheme, weight_db_logic)
dosage_loader = GWASUtilities.GWASDosageFileLoader(input_path, self.compressed, self.args.separator, callback)
result_sets = dosage_loader.load()
# The following check is sort of redundant, as it exists in "saveSetsToCompressedFile".
# It exists merely to provide different login
if len(result_sets):
KeyedDataSet.KeyedDataSetFileUtilities.saveSetsToCompressedFile(output_path, result_sets, "rsid")
else:
logging.info("No snps from the tissue model found in the GWAS file")
