def run_generate(self, args):
self._check_generate(args)
generate_model = GenerateModel()
generate_model.generate_pipeline(args.input_matrix, args.groups,
args.model_type, args.testing_portion,
args.grid_search, args.threads,
args.output)
def test_hello_generate(self):
tmp = tempfile.mkdtemp()
generate_model = GenerateModel()
generate_model.generate_pipeline(
self.sample_matrix_path,
self.sample_metadata_path,
generate_model.classifier,
0.2, # Default testing portion
False, # Dont do a grid search for fine tuning
2, # Threads
tmp # Output directory
)
expected_files = sorted(os.listdir(tmp))
observed_files = sorted(os.listdir(self.sample_generate_path))
self.assertEqual(len(expected_files), len(observed_files))
for expected_file, observed_file in zip(expected_files,
observed_files):
#expected_file_path = os.path.join(tmp, expected_file)
#observed_file_path = os.path.join(self.sample_generate_path, observed_file)
# Are all files present?
self.assertEqual(expected_file, observed_file)
shutil.rmtree(tmp)
def do(self, forester_model_directory, input_matrix_path,
output_directory):
'''
Inputs
------
Outputs
-------
'''
forester_model = self.parse_input_model_directory(
forester_model_directory)
logging.info('Parsing input')
gm = GenerateModel()
logging.info('Loading model: %s' % (self.RF_MODEL))
logging.info('Parsing data')
features, attribute_list \
= gm.parse_input_matrix(input_matrix_path)
sample_list = list()
content_list = list()
for sample, content in features.items():
sample_list.append(sample)
sample_content = []
for attribute in forester_model['attribute_list.txt']:
if attribute in content:
sample_content.append(content[attribute])
else:
sample_content.append('0')
content_list.append(sample_content)
logging.info('Making predictions')
output_lines = self._make_predictions(forester_model[self.RF_MODEL],
sample_list, content_list,
forester_model[self.LABELS_DICT])
self._write_predictions(output_lines, output_directory)
def run_enrichm(self, args, command):
'''
Parameters
----------
Output
------
'''
self._check_general(args)
self._logging_setup(args)
logging.info("Command: %s" % ' '.join(command))
logging.info("Running the %s pipeline" % args.subparser_name)
if args.subparser_name == self.DATA:
d = Data()
d.do(args.uninstall, args.dry)
if args.subparser_name == self.ANNOTATE:
self._check_annotate(args)
annotate = Annotate(# Define inputs and outputs
args.output,
# Define type of annotation to be carried out
args.ko, args.ko_hmm, args.pfam, args.tigrfam,
args.clusters, args.orthologs, args.cazy,
args.ec,
# Cutoffs
args.evalue, args.bit, args.id, args.aln_query,
args.aln_reference, args.c, args.cut_ga,
args.cut_nc, args.cut_tc, args.cut_ko,
args.inflation, args.chunk_number, args.chunk_max,
args.count_domains,
# Parameters
args.threads, args.parallel, args.suffix, args.light)
annotate.annotate_pipeline(args.genome_directory,
args.protein_directory,
args.genome_files,
args.protein_files)
elif args.subparser_name == self.CLASSIFY:
self._check_classify(args)
classify = Classify()
classify.classify_pipeline(args.custom_modules, args.cutoff, args.aggregate,
args.genome_and_annotation_matrix, args.output)
elif args.subparser_name == self.ENRICHMENT:
self._check_enrichment(args)
enrichment = Enrichment()
enrichment.enrichment_pipeline(# Input options
args.annotate_output, args.annotation_matrix,
args.metadata, args.abundance, args.abundance_metadata,
args.transcriptome, args.transcriptome_metadata,
# Runtime options
args.pval_cutoff, args.proportions_cutoff,
args.threshold, args.multi_test_correction,
args.batchfile, args.processes,
args.allow_negative_values, args.ko, args.pfam,
args.tigrfam, args.cluster, args.ortholog, args.cazy,
args.ec, args.ko_hmm,
# Outputs
args.output)
elif(args.subparser_name == NetworkAnalyser.PATHWAY or
args.subparser_name == NetworkAnalyser.EXPLORE):
self._check_network(args)
network_analyser=NetworkAnalyser()
network_analyser.network_pipeline(args.subparser_name, args.matrix,
args.genome_metadata, args.tpm_values,
args.tpm_metadata, args.abundance,
args.abundance_metadata, args.metabolome,
args.enrichment_output, args.depth, args.filter,
args.limit, args.queries, args.output)
if args.subparser_name == self.PREDICT:
self._check_predict(args)
predict = Predict()
predict.predict_pipeline(args.forester_model_directory,
args.input_matrix,
args.output)
elif args.subparser_name == self.GENERATE:
self._check_generate(args)
generate_model = GenerateModel()
generate_model.generate_pipeline(args.input_matrix,
args.groups,
args.model_type,
args.testing_portion,
args.grid_search,
args.threads,
args.output)
elif args.subparser_name == self.USES:
self._check_uses(args)
uses = Uses()
uses.uses_pipeline(args.compounds_list,
args.annotation_matrix,
args.metadata,
args.output,
args.count)
logging.info('Finished running EnrichM')
def main(self, args, command):
'''
Parameters
----------
Output
------
'''
self._check_general(args)
self._logging_setup(args)
logging.info("Running command: %s" % ' '.join(command))
if args.subparser_name == self.DATA:
d = Data()
d.do(args.uninstall)
if args.subparser_name == self.ANNOTATE:
self._check_annotate(args)
a = Annotate(# Define inputs and outputs
args.output,
# Define type of annotation to be carried out
args.ko,
args.pfam,
args.tigrfam,
args.hypothetical,
args.cazy,
# Cutoffs
args.evalue,
args.bit,
args.id,
args.aln_query,
args.aln_reference,
args.c,
args.cut_ga,
args.cut_nc,
args.cut_tc,
args.inflation,
args.chunk_number,
args.chunk_max,
# Parameters
args.threads,
args.parallel,
args.suffix,
args.light)
a.do(args.genome_directory,
args.protein_directory,
args.genome_files,
args.protein_files)
elif args.subparser_name == self.CLASSIFY:
self._check_classify(args)
c = Classify()
c.do(args.custom_modules,
args.cutoff,
args.genome_and_annotation_file,
args.genome_and_annotation_matrix,
args.output)
elif args.subparser_name == self.ENRICHMENT:
self._check_enrichment(args)
e = Enrichment()
e.do(# Input options
args.annotate_output,
args.metadata,
args.modules,
args.abundances,
# Runtime options
args.genomes_to_compare_with_group,
args.pval_cutoff,
args.proportions_cutoff,
args.threshold,
args.multi_test_correction,
args.batchfile,
args.processes,
args.ko,
args.pfam,
args.tigrfam,
args.hypothetical,
args.cazy,
# Outputs
args.output)
elif args.subparser_name == self.CONNECT:
self._check_connect(args)
c = Connect()
c.do(args.annotate_output,
args.metadata,
args.custom_modules,
args.cutoff,
args.output)
elif(args.subparser_name == NetworkAnalyser.PATHWAY or
args.subparser_name == NetworkAnalyser.EXPLORE or
args.subparser_name == NetworkAnalyser.TRAVERSE):
self._check_network(args)
na=NetworkAnalyser(args.metadata)
na.do(args.matrix,
args.transcriptome,
args.metabolome,
args.depth,
args.filter,
args.limit,
args.queries,
args.subparser_name,
args.starting_compounds,
args.steps,
args.number_of_queries,
args.output)
if args.subparser_name == self.PREDICT:
self._check_predict(args)
p = Predict()
p.do(args.forester_model_directory,
args.input_matrix,
args.output)
elif args.subparser_name == self.GENERATE:
self._check_generate(args)
gm = GenerateModel()
gm.do(args.input_matrix,
args.groups,
args.model_type,
args.testing_portion,
args.grid_search,
args.threads,
args.output)
logging.info('Done!')