scg

#!/usr/bin/env python

#=======================================================================================================================
# Single Cell Genotyper
# Author : Andrew Roth
#=======================================================================================================================
from scg.run import run_doublet_analysis, run_singlet_analysis
from scg.run_dmm import run_dirichlet_mixture_model_analysis

import argparse

parser = argparse.ArgumentParser(prog='Single Cell Genotyper')

parser.add_argument('--version', action='version', version='0.3.1')

subparsers = parser.add_subparsers()

#=======================================================================================================================
# Shared analysis parser
#=======================================================================================================================
analysis_parser = argparse.ArgumentParser(add_help=False)

analysis_parser.add_argument('--config_file', required=True,
                            help='''Path to YAML format configuration file.''')

analysis_parser.add_argument('--lower_bound_file', default=None,
                             help='''Path of file where lower bound will be written.''')

analysis_parser.add_argument('--out_dir', default=None,
                            help='''Path where output files will be written.''')

analysis_parser.add_argument('--convergence_tolerance', default=1e-4, type=float)

analysis_parser.add_argument('--max_num_iters', default=1000, type=int)

analysis_parser.add_argument('--seed', default=None, type=int,
                             help='''Set random seed so results can be reproduced. By default a random seed is
                             chosen.''')

analysis_parser.add_argument('--labels_file', default=None,
                             help='''Path of file with initial labels to use.''')

#---------------------------------------------------------------------------------------------------------------------- 
genotyper_parser = argparse.ArgumentParser(add_help=False)

genotyper_parser.add_argument('--use_position_specific_error_rate', action='store_true', default=False,
                              help='''If an error rate will be estimated for each position.''')

genotyper_parser.add_argument('--samples_file', default=None,
                              help='''Path mapping cells to samples. If set each sample will have a separate mixing
                              proportion.''')

#---------------------------------------------------------------------------------------------------------------------- 
doublet_parser = subparsers.add_parser('run_doublet_model', parents=[analysis_parser, genotyper_parser],
                                       help='''Analyse single cell data using the single cell genotyper model accounting for doublets.''')

doublet_parser.add_argument('--state_map_file', required=True)

doublet_parser.set_defaults(func=run_doublet_analysis)

#---------------------------------------------------------------------------------------------------------------------- 
singlet_parser = subparsers.add_parser('run_singlet_model', parents=[analysis_parser, genotyper_parser],
                                       help='''Analyse single cell data using the single cell genotyper model.''')

singlet_parser.set_defaults(func=run_singlet_analysis)

#---------------------------------------------------------------------------------------------------------------------- 
dmm_parser = subparsers.add_parser('run_dirichlet_mixture_model', parents=[analysis_parser],
                                   help='''Analyse single cell data using a mixture of Dirichlet distributions.''')

dmm_parser.set_defaults(func=run_dirichlet_mixture_model_analysis)

#---------------------------------------------------------------------------------------------------------------------- 
args = parser.parse_args()

args.func(args)