def _build_mutations_file(in_file, out_file, prior):
config = {}
reader = csv.DictReader(open(in_file), delimiter='\t')
config['mutations'] = []
for row in reader:
mutation_id = row['mutation_id']
ref_counts = int(row['ref_counts'])
var_counts = int(row['var_counts'])
normal_cn = int(row['normal_cn'])
minor_cn = int(row['minor_cn'])
major_cn = int(row['major_cn'])
mutation = get_mutation(mutation_id, ref_counts, var_counts, normal_cn,
minor_cn, major_cn, prior)
config['mutations'].append(mutation.to_dict())
make_parent_directory(out_file)
fh = open(out_file, 'w')
yaml.dump(config, fh, Dumper=Dumper)
fh.close()
def run_analysis_pipeline(args):
config_file = _setup_analysis(
density=args.density,
in_files=args.in_files,
init_method=args.init_method,
num_iters=args.num_iters,
samples=args.samples,
prior=args.prior,
tumour_contents=args.tumour_contents,
working_dir=args.working_dir,
config_extras_file=args.config_extras_file,
)
_run_analysis(config_file, args.seed)
tables_dir = os.path.join(args.working_dir, 'tables')
make_directory(tables_dir)
for table_type in ['cluster', 'loci']:
out_file = os.path.join(tables_dir, '{0}.tsv'.format(table_type))
_build_table(config_file=config_file,
out_file=out_file,
burnin=args.burnin,
max_clusters=args.max_clusters,
mesh_size=args.mesh_size,
table_type=table_type,
thin=args.thin)
plots_dir = os.path.join(args.working_dir, 'plots')
plots = [('cluster', 'density'), ('cluster', 'parallel_coordinates'),
('cluster', 'scatter'), ('loci', 'density'),
('loci', 'parallel_coordinates'), ('loci', 'scatter'),
('loci', 'similarity_matrix'),
('loci', 'vaf_parallel_coordinates'), ('loci', 'vaf_scatter')]
for category, plot_type in plots:
plot_file = os.path.join(
plots_dir, category, '{0}.{1}'.format(plot_type,
args.plot_file_format))
make_parent_directory(plot_file)
if category == 'cluster':
_cluster_plot(config_file, plot_file, args.burnin,
args.max_clusters, args.mesh_size,
args.min_cluster_size, plot_type, args.samples,
args.thin)
elif category == 'loci':
_loci_plot(config_file,
plot_file,
plot_type,
burnin=args.burnin,
min_cluster_size=args.min_cluster_size,
samples=args.samples,
thin=args.thin)