Mercurial > repos > greg > multigps
comparison multigps.py @ 5:7b15099378a7 draft
Uploaded
author | greg |
---|---|
date | Sat, 02 Jan 2016 14:07:37 -0500 |
parents | |
children | afba8c2a3b32 |
comparison
equal
deleted
inserted
replaced
4:32cff4d1d0ff | 5:7b15099378a7 |
---|---|
1 import argparse | |
2 import os | |
3 import shutil | |
4 import subprocess | |
5 import sys | |
6 import tempfile | |
7 | |
8 BUFF_SIZE = 1048576 | |
9 DESIGN_FILE = 'design.tabular' | |
10 | |
11 | |
12 def generate_design_file(input_items, filename): | |
13 design_file = open(filename, 'wb') | |
14 for items in input_items: | |
15 filename, ext, signal_control, condition_name, replicate_name, read_distribution_file, fixed_read = items | |
16 # MultiGPS version 0.5 does not support the newer scidx datatype. | |
17 if ext == 'scidx': | |
18 datatype = 'IDX' | |
19 else: | |
20 datatype = ext.upper() | |
21 design_file.write('%s\n' % '\t'.join([filename, signal_control, datatype, condition_name, replicate_name, read_distribution_file, fixed_read])) | |
22 design_file.close() | |
23 | |
24 | |
25 def get_stderr_exception(tmp_err, tmp_stderr, tmp_out, tmp_stdout, include_stdout=False): | |
26 tmp_stderr.close() | |
27 # Get stderr, allowing for case where it's very large. | |
28 tmp_stderr = open(tmp_err, 'rb') | |
29 stderr_str = '' | |
30 buffsize = BUFF_SIZE | |
31 try: | |
32 while True: | |
33 stderr_str += tmp_stderr.read(buffsize) | |
34 if not stderr_str or len(stderr_str) % buffsize != 0: | |
35 break | |
36 except OverflowError: | |
37 pass | |
38 tmp_stderr.close() | |
39 if include_stdout: | |
40 tmp_stdout = open(tmp_out, 'rb') | |
41 stdout_str = '' | |
42 buffsize = BUFF_SIZE | |
43 try: | |
44 while True: | |
45 stdout_str += tmp_stdout.read(buffsize) | |
46 if not stdout_str or len(stdout_str) % buffsize != 0: | |
47 break | |
48 except OverflowError: | |
49 pass | |
50 tmp_stdout.close() | |
51 if include_stdout: | |
52 return 'STDOUT\n%s\n\nSTDERR\n%s\n' % (stdout_str, stderr_str) | |
53 return stderr_str | |
54 | |
55 | |
56 def stop_err(msg): | |
57 sys.stderr.write(msg) | |
58 sys.exit(1) | |
59 | |
60 parser = argparse.ArgumentParser() | |
61 parser.add_argument('--input_item', dest='input_items', action='append', nargs=7, help="Input datasets and associated parameters") | |
62 parser.add_argument('--threads', dest='threads', type=int, default=4, help='The number of threads to run') | |
63 parser.add_argument('--multigps_jar', dest='multigps_jar', help='Path to the MultiGPS jar file') | |
64 parser.add_argument('--geninfo', dest='geninfo', help='File listing the lengths of all chromosomes') | |
65 parser.add_argument('--use_motif', dest='use_motif', help='Perform motif-finding or use a motif-prior') | |
66 parser.add_argument('--seq_dir', dest='seq_dir', default=None, help='Directory containing fasta files corresponding to every named chromosome') | |
67 parser.add_argument('--positional_prior', dest='positional_prior', default=None, help='Perform inter-experiment positional prior') | |
68 parser.add_argument('--events_shared_probability', dest='events_shared_probability', type=float, default=0.9, help='Probability that events are shared across conditions') | |
69 parser.add_argument('--motifs', dest='motifs', default=None, help='Perform motif-finding and motif priors') | |
70 parser.add_argument('--motif_finding_only', dest='motif_finding_only', default=None, help='Perform motif-finding only') | |
71 parser.add_argument('--num_motifs', dest='num_motifs', type=int, default=None, help='Number of motifs MEME should find for each condition') | |
72 parser.add_argument('--mememinw', dest='mememinw', type=int, default=None, help='Minimum motif width for MEME') | |
73 parser.add_argument('--mememaxw', dest='mememaxw', type=int, default=None, help='Maximum motif width for MEME') | |
74 parser.add_argument('--fixedpb', dest='fixedpb', type=int, default=None, help='Fixed per-base limit') | |
75 parser.add_argument('--poissongausspb', dest='poissongausspb', type=int, default=None, help='Poisson threshold for filtering per base') | |
76 parser.add_argument('--non_unique_reads', dest='non_unique_reads', default=None, help='Use non-unique reads') | |
77 parser.add_argument('--minqvalue', dest='minqvalue', type=int, default=None, help='Minimum Q-value (corrected p-value) of reported binding events') | |
78 parser.add_argument('--minfold', dest='minfold', type=int, default=None, help='Minimum event fold-change vs scaled control') | |
79 parser.add_argument('--diff_enrichment_tests', dest='diff_enrichment_tests', help='Run differential enrichment tests') | |
80 parser.add_argument('--edgerod', dest='edgerod', type=int, default=None, help='EdgeR over-dispersion parameter value') | |
81 parser.add_argument('--diffp', dest='diffp', type=int, default=None, help='Minimum p-value for reporting differential enrichment') | |
82 parser.add_argument('--noscaling', dest='noscaling', default=None, help='Do not use signal vs control scaling') | |
83 parser.add_argument('--medianscale', dest='medianscale', default=None, help='Use the median signal/control ratio as the scaling factor') | |
84 parser.add_argument('--sesscale', dest='sesscale', default=None, help='Estimate scaling factor by SES') | |
85 parser.add_argument('--scalewin', dest='scalewin', type=int, default=None, help='Window size for estimating scaling ratios') | |
86 parser.add_argument('--max_training_rounds', dest='max_training_rounds', type=int, default=None, help='Maximum number of training rounds for updating binding event read distributions') | |
87 parser.add_argument('--exclude_file', dest='exclude_file', default=None, help='File containing a set of regions to ignore during MultiGPS training') | |
88 parser.add_argument('--binding_model_updates', dest='binding_model_updates', default=None, help='Perform binding model updates') | |
89 parser.add_argument('--minmodelupdateevents', dest='minmodelupdateevents', type=int, default=None, help='Minimum number of events to support an update of the read distribution') | |
90 parser.add_argument('--binding_model_smoothing', dest='binding_model_smoothing', default=None, help='Binding model smoothing') | |
91 parser.add_argument('--spline_smooth', dest='spline_smooth', type=int, default=None, help='Binding model smoothing value') | |
92 parser.add_argument('--gauss_smooth', dest='gauss_smooth', type=int, default=None, help='Gaussian smoothing std dev') | |
93 parser.add_argument('--joint_in_model', dest='joint_in_model', default=None, help='Allow joint events in model updates') | |
94 parser.add_argument('--ml_config_not_shared', dest='ml_config_not_shared', default=None, help='Share component configs in the ML step') | |
95 parser.add_argument('--output_html_path', dest='output_html_path', help='Output html results file') | |
96 parser.add_argument('--output_html_files_path', dest='output_html_files_path', help='Output html extra files') | |
97 parser.add_argument('--output_process_path', dest='output_process_path', default=None, help='Output file for capturing stdout') | |
98 args = parser.parse_args() | |
99 | |
100 try: | |
101 # Preparation. | |
102 tmp_dir = tempfile.mkdtemp(prefix="tmp-multigps-") | |
103 generate_design_file(args.input_items, DESIGN_FILE) | |
104 # Build the command line. | |
105 cmd_list = ['java -jar %s --verbose' % args.multigps_jar] | |
106 cmd_list.append('--threads %d' % args.threads) | |
107 cmd_list.append('--geninfo %s' % args.geninfo) | |
108 if args.use_motif == 'yes': | |
109 cmd_list.append('--seq %s' % args.seq_dir) | |
110 if args.positional_prior == 'no': | |
111 cmd_list.append('--noposprior') | |
112 cmd_list.append('--probshared %3f' % args.events_shared_probability) | |
113 if args.motifs == 'no': | |
114 cmd_list.append('--nomotifs') | |
115 if args.motif_finding_only == 'yes': | |
116 cmd_list.append('--nomotifprior') | |
117 if args.num_motifs is not None: | |
118 cmd_list.append('--memenmotifs %d' % args.num_motifs) | |
119 if args.mememinw is not None: | |
120 cmd_list.append('--mememinw %d' % args.mememinw) | |
121 if args.mememaxw is not None: | |
122 cmd_list.append('--mememaxw %d' % args.mememaxw) | |
123 if args.fixedpb is not None: | |
124 cmd_list.append('--fixedpb %d' % args.fixedpb) | |
125 if args.poissongausspb is not None: | |
126 cmd_list.append('--poissongausspb %d' % args.poissongausspb) | |
127 if args.non_unique_reads == 'yes': | |
128 cmd_list.append('--nonunique') | |
129 if args.minqvalue is not None: | |
130 cmd_list.append('--q %d' % args.minqvalue) | |
131 if args.minfold is not None: | |
132 cmd_list.append('--minfold %d' % args.minfold) | |
133 if args.diff_enrichment_tests == 'no': | |
134 cmd_list.append('--nodifftests') | |
135 if args.edgerod is not None: | |
136 cmd_list.append('--edgerod %d' % args.edgerod) | |
137 if args.diffp is not None: | |
138 cmd_list.append('--diffp %d' % args.diffp) | |
139 if args.noscaling == 'no': | |
140 cmd_list.append('--noscaling') | |
141 if args.medianscale == "yes": | |
142 cmd_list.append('--medianscale') | |
143 if args.sesscale == "yes": | |
144 cmd_list.append('--sesscale') | |
145 if args.scalewin is not None: | |
146 cmd_list.append('--scalewin %d' % args.scalewin) | |
147 if args.max_training_rounds is not None: | |
148 cmd_list.append('--r %d' % args.max_training_rounds) | |
149 if args.exclude_file not in [None, 'None']: | |
150 cmd_list.append('--exclude %s' % args.exclude_file) | |
151 if args.binding_model_updates == 'no': | |
152 cmd_list.append('--nomodelupdate') | |
153 if args.minmodelupdateevents is not None: | |
154 cmd_list.append('--minmodelupdateevents %d' % args.minmodelupdateevents) | |
155 if args.binding_model_smoothing == 'yes': | |
156 if args.spline_smooth is not None: | |
157 cmd_list.append('--splinesmoothparam %d' % args.spline_smooth) | |
158 else: | |
159 cmd_list.append('--nomodelsmoothing') | |
160 if args.gauss_smooth is not None: | |
161 cmd_list.append('--gaussmodelsmoothing') | |
162 cmd_list.append('--gausssmoothparam %d' % args.gauss_smooth) | |
163 if args.joint_in_model == 'yes': | |
164 cmd_list.append('--jointinmodel') | |
165 if args.ml_config_not_shared == 'no': | |
166 cmd_list.append('--mlconfignotshared') | |
167 cmd_list.append('--design %s' % DESIGN_FILE) | |
168 cmd_list.append('--out %s' % args.output_html_files_path) | |
169 if args.output_process_path is None: | |
170 cmd_list.append('2>/dev/null') | |
171 else: | |
172 cmd_list.append('2>%s' % args.output_process_path) | |
173 cmd = ' '.join(cmd_list) | |
174 # Define command response buffers. | |
175 tmp_out = tempfile.NamedTemporaryFile(dir=tmp_dir).name | |
176 tmp_stdout = open(tmp_out, 'wb') | |
177 tmp_err = tempfile.NamedTemporaryFile(dir=tmp_dir).name | |
178 tmp_stderr = open(tmp_err, 'wb') | |
179 tmp_filename = tempfile.NamedTemporaryFile(dir=tmp_dir).name | |
180 # Execute the command. | |
181 proc = subprocess.Popen(args=cmd, stderr=tmp_stderr, stdout=tmp_stdout, shell=True) | |
182 rc = proc.wait() | |
183 if rc != 0: | |
184 error_message = get_stderr_exception(tmp_err, tmp_stderr, tmp_out, tmp_stdout) | |
185 stop_err(error_message) | |
186 # Move the output HTML file to the output dataset path. | |
187 output_html_filename = 'multiGPS_%s_results.html' % os.path.split(args.output_html_files_path)[1] | |
188 shutil.move(os.path.join(args.output_html_files_path, output_html_filename), args.output_html_path) | |
189 except Exception, e: | |
190 stop_err('Error running MultiGPS\n%s\n' % str(e)) | |
191 finally: | |
192 if os.path.exists(tmp_dir): | |
193 shutil.rmtree(tmp_dir) |