Mercurial > repos > greg > vsnp_determine_ref_from_data

comparison vsnp_determine_ref_from_data.py @ 0:ebc08e5ce646 draft

Find changesets by keywords (author, files, the commit message), revision number or hash, or revset expression.
Uploaded
author greg
date Tue, 21 Apr 2020 10:08:28 -0400
parents
children 36bdf8b439ed
comparison
equal deleted inserted replaced
-1:000000000000 0:ebc08e5ce646
1 #!/usr/bin/env python
2
3 import argparse
4 import gzip
5 import multiprocessing
6 import os
7 import queue
8 import yaml
9 from Bio.SeqIO.QualityIO import FastqGeneralIterator
10 from collections import OrderedDict
11
12 INPUT_READS_DIR = 'input_reads'
13 OUTPUT_DBKEY_DIR = 'output_dbkey'
14 OUTPUT_METRICS_DIR = 'output_metrics'
15
16
17 def get_base_file_name(file_path):
18 base_file_name = os.path.basename(file_path)
19 if base_file_name.find(".") > 0:
20 # Eliminate the extension.
21 return os.path.splitext(base_file_name)[0]
22 elif base_file_name.find("_") > 0:
23 # The dot extension was likely changed to
24 # the " character.
25 items = base_file_name.split("_")
26 no_ext = "_".join(items[0:-2])
27 if len(no_ext) > 0:
28 return no_ext
29 return base_file_name
30
31
32 def get_dbkey(dnaprints_dict, key, s):
33 # dnaprints_dict looks something like this:
34 # {'brucella': {'NC_002945v4': ['11001110', '11011110', '11001100']}
35 # {'bovis': {'NC_006895': ['11111110', '00010010', '01111011']}}
36 d = dnaprints_dict.get(key, {})
37 for data_table_value, v_list in d.items():
38 if s in v_list:
39 return data_table_value
40 return ""
41
42
43 def get_dnaprints_dict(dnaprint_fields):
44 # A dndprint_fields entry looks something liek this.
45 # [['AF2122', '/galaxy/tool-data/vsnp/AF2122/dnaprints/NC_002945v4.yml']]
46 dnaprints_dict = {}
47 for item in dnaprint_fields:
48 # Here item is a 2-element list of data
49 # table components, # value and path.
50 value = item[0]
51 path = item[1].strip()
52 with open(path, "rt") as fh:
53 # The format of all dnaprints yaml
54 # files is something like this:
55 # brucella:
56 # - 0111111111111111
57 print_dict = yaml.load(fh, Loader=yaml.Loader)
58 for print_dict_k, print_dict_v in print_dict.items():
59 dnaprints_v_dict = dnaprints_dict.get(print_dict_k, {})
60 if len(dnaprints_v_dict) > 0:
61 # dnaprints_dict already contains k (e.g., 'brucella',
62 # and dnaprints_v_dict will be a dictionary # that
63 # looks something like this:
64 # {'NC_002945v4': ['11001110', '11011110', '11001100']}
65 value_list = dnaprints_v_dict.get(value, [])
66 value_list = value_list + print_dict_v
67 dnaprints_v_dict[value] = value_list
68 else:
69 # dnaprints_v_dict is an empty dictionary.
70 dnaprints_v_dict[value] = print_dict_v
71 dnaprints_dict[print_dict_k] = dnaprints_v_dict
72 # dnaprints_dict looks something like this:
73 # {'brucella': {'NC_002945v4': ['11001110', '11011110', '11001100']}
74 # {'bovis': {'NC_006895': ['11111110', '00010010', '01111011']}}
75 return dnaprints_dict
76
77
78 def get_group_and_dbkey(dnaprints_dict, brucella_string, brucella_sum, bovis_string, bovis_sum, para_string, para_sum):
79 if brucella_sum > 3:
80 group = "Brucella"
81 dbkey = get_dbkey(dnaprints_dict, "brucella", brucella_string)
82 elif bovis_sum > 3:
83 group = "TB"
84 dbkey = get_dbkey(dnaprints_dict, "bovis", bovis_string)
85 elif para_sum >= 1:
86 group = "paraTB"
87 dbkey = get_dbkey(dnaprints_dict, "para", para_string)
88 else:
89 group = ""
90 dbkey = ""
91 return group, dbkey
92
93
94 def get_group_and_dbkey_for_collection(task_queue, finished_queue, dnaprints_dict, timeout):
95 while True:
96 try:
97 tup = task_queue.get(block=True, timeout=timeout)
98 except queue.Empty:
99 break
100 fastq_file, count_list, brucella_string, brucella_sum, bovis_string, bovis_sum, para_string, para_sum = tup
101 group, dbkey = get_group_and_dbkey(dnaprints_dict, brucella_string, brucella_sum, bovis_string, bovis_sum, para_string, para_sum)
102 finished_queue.put((fastq_file, count_list, group, dbkey))
103 task_queue.task_done()
104
105
106 def get_oligo_dict():
107 oligo_dict = {}
108 oligo_dict["01_ab1"] = "AATTGTCGGATAGCCTGGCGATAACGACGC"
109 oligo_dict["02_ab3"] = "CACACGCGGGCCGGAACTGCCGCAAATGAC"
110 oligo_dict["03_ab5"] = "GCTGAAGCGGCAGACCGGCAGAACGAATAT"
111 oligo_dict["04_mel"] = "TGTCGCGCGTCAAGCGGCGTGAAATCTCTG"
112 oligo_dict["05_suis1"] = "TGCGTTGCCGTGAAGCTTAATTCGGCTGAT"
113 oligo_dict["06_suis2"] = "GGCAATCATGCGCAGGGCTTTGCATTCGTC"
114 oligo_dict["07_suis3"] = "CAAGGCAGATGCACATAATCCGGCGACCCG"
115 oligo_dict["08_ceti1"] = "GTGAATATAGGGTGAATTGATCTTCAGCCG"
116 oligo_dict["09_ceti2"] = "TTACAAGCAGGCCTATGAGCGCGGCGTGAA"
117 oligo_dict["10_canis4"] = "CTGCTACATAAAGCACCCGGCGACCGAGTT"
118 oligo_dict["11_canis"] = "ATCGTTTTGCGGCATATCGCTGACCACAGC"
119 oligo_dict["12_ovis"] = "CACTCAATCTTCTCTACGGGCGTGGTATCC"
120 oligo_dict["13_ether2"] = "CGAAATCGTGGTGAAGGACGGGACCGAACC"
121 oligo_dict["14_63B1"] = "CCTGTTTAAAAGAATCGTCGGAACCGCTCT"
122 oligo_dict["15_16M0"] = "TCCCGCCGCCATGCCGCCGAAAGTCGCCGT"
123 oligo_dict["16_mel1b"] = "TCTGTCCAAACCCCGTGACCGAACAATAGA"
124 oligo_dict["17_tb157"] = "CTCTTCGTATACCGTTCCGTCGTCACCATGGTCCT"
125 oligo_dict["18_tb7"] = "TCACGCAGCCAACGATATTCGTGTACCGCGACGGT"
126 oligo_dict["19_tbbov"] = "CTGGGCGACCCGGCCGACCTGCACACCGCGCATCA"
127 oligo_dict["20_tb5"] = "CCGTGGTGGCGTATCGGGCCCCTGGATCGCGCCCT"
128 oligo_dict["21_tb2"] = "ATGTCTGCGTAAAGAAGTTCCATGTCCGGGAAGTA"
129 oligo_dict["22_tb3"] = "GAAGACCTTGATGCCGATCTGGGTGTCGATCTTGA"
130 oligo_dict["23_tb4"] = "CGGTGTTGAAGGGTCCCCCGTTCCAGAAGCCGGTG"
131 oligo_dict["24_tb6"] = "ACGGTGATTCGGGTGGTCGACACCGATGGTTCAGA"
132 oligo_dict["25_para"] = "CCTTTCTTGAAGGGTGTTCG"
133 oligo_dict["26_para_sheep"] = "CGTGGTGGCGACGGCGGCGGGCCTGTCTAT"
134 oligo_dict["27_para_cattle"] = "TCTCCTCGGTCGGTGATTCGGGGGCGCGGT"
135 return oligo_dict
136
137
138 def get_seq_counts(value, fastq_list, gzipped):
139 count = 0
140 for fastq_file in fastq_list:
141 if gzipped == "true":
142 with gzip.open(fastq_file, 'rt') as fh:
143 for title, seq, qual in FastqGeneralIterator(fh):
144 count += seq.count(value)
145 else:
146 with open(fastq_file, 'r') as fh:
147 for title, seq, qual in FastqGeneralIterator(fh):
148 count += seq.count(value)
149 return(value, count)
150
151
152 def get_species_counts(fastq_list, gzipped):
153 count_summary = {}
154 oligo_dict = get_oligo_dict()
155 for v1 in oligo_dict.values():
156 returned_value, count = get_seq_counts(v1, fastq_list, gzipped)
157 for key, v2 in oligo_dict.items():
158 if returned_value == v2:
159 count_summary.update({key: count})
160 count_list = []
161 for v in count_summary.values():
162 count_list.append(v)
163 brucella_sum = sum(count_list[:16])
164 bovis_sum = sum(count_list[16:24])
165 para_sum = sum(count_list[24:])
166 return count_summary, count_list, brucella_sum, bovis_sum, para_sum
167
168
169 def get_species_counts_for_collection(task_queue, finished_queue, gzipped, timeout):
170 while True:
171 try:
172 fastq_file = task_queue.get(block=True, timeout=timeout)
173 except queue.Empty:
174 break
175 count_summary, count_list, brucella_sum, bovis_sum, para_sum = get_species_counts([fastq_file], gzipped)
176 finished_queue.put((fastq_file, count_summary, count_list, brucella_sum, bovis_sum, para_sum))
177 task_queue.task_done()
178
179
180 def get_species_strings(count_summary):
181 binary_dictionary = {}
182 for k, v in count_summary.items():
183 if v > 1:
184 binary_dictionary.update({k: 1})
185 else:
186 binary_dictionary.update({k: 0})
187 binary_dictionary = OrderedDict(sorted(binary_dictionary.items()))
188 binary_list = []
189 for v in binary_dictionary.values():
190 binary_list.append(v)
191 brucella_binary = binary_list[:16]
192 brucella_string = ''.join(str(e) for e in brucella_binary)
193 bovis_binary = binary_list[16:24]
194 bovis_string = ''.join(str(e) for e in bovis_binary)
195 para_binary = binary_list[24:]
196 para_string = ''.join(str(e) for e in para_binary)
197 return brucella_string, bovis_string, para_string
198
199
200 def get_species_strings_for_collection(task_queue, finished_queue, timeout):
201 while True:
202 try:
203 tup = task_queue.get(block=True, timeout=timeout)
204 except queue.Empty:
205 break
206 fastq_file, count_summary, count_list, brucella_sum, bovis_sum, para_sum = tup
207 brucella_string, bovis_string, para_string = get_species_strings(count_summary)
208 finished_queue.put((fastq_file, count_list, brucella_string, brucella_sum, bovis_string, bovis_sum, para_string, para_sum))
209 task_queue.task_done()
210
211
212 def output_dbkey(file_name, dbkey, output_file=None):
213 # Output the dbkey.
214 if output_file is None:
215 # We're producing a dataset collection.
216 output_file = os.path.join(OUTPUT_DBKEY_DIR, "%s.txt" % file_name)
217 with open(output_file, "w") as fh:
218 fh.write("%s" % dbkey)
219
220
221 def output_files(fastq_file, count_list, group, dbkey, dbkey_file=None, metrics_file=None):
222 base_file_name = get_base_file_name(fastq_file)
223 if dbkey_file is not None:
224 # We're dealing with a single read or
225 # a set of paired reads. If the latter,
226 # the following will hopefully produce a
227 # good sample string.
228 if base_file_name.find("_") > 0:
229 base_file_name = base_file_name.split("_")[0]
230 output_dbkey(base_file_name, dbkey, dbkey_file)
231 output_metrics(base_file_name, count_list, group, dbkey, metrics_file)
232
233
234 def output_files_for_collection(task_queue, timeout):
235 while True:
236 try:
237 tup = task_queue.get(block=True, timeout=timeout)
238 except queue.Empty:
239 break
240 fastq_file, count_list, group, dbkey = tup
241 output_files(fastq_file, count_list, group, dbkey)
242 task_queue.task_done()
243
244
245 def output_metrics(file_name, count_list, group, dbkey, output_file=None):
246 # Output the metrics.
247 if output_file is None:
248 # We're producing a dataset collection.
249 output_file = os.path.join(OUTPUT_METRICS_DIR, "%s.txt" % file_name)
250 with open(output_file, "w") as fh:
251 fh.write("Sample: %s\n" % file_name)
252 fh.write("Brucella counts: ")
253 for i in count_list[:16]:
254 fh.write("%d," % i)
255 fh.write("\nTB counts: ")
256 for i in count_list[16:24]:
257 fh.write("%d," % i)
258 fh.write("\nPara counts: ")
259 for i in count_list[24:]:
260 fh.write("%d," % i)
261 fh.write("\nGroup: %s" % group)
262 fh.write("\ndbkey: %s\n" % dbkey)
263
264
265 def set_num_cpus(num_files, processes):
266 num_cpus = int(multiprocessing.cpu_count())
267 if num_files < num_cpus and num_files < processes:
268 return num_files
269 if num_cpus < processes:
270 half_cpus = int(num_cpus / 2)
271 if num_files < half_cpus:
272 return num_files
273 return half_cpus
274 return processes
275
276
277 if __name__ == '__main__':
278 parser = argparse.ArgumentParser()
279
280 parser.add_argument('--dnaprint_fields', action='append', dest='dnaprint_fields', nargs=2, help="List of dnaprints data table value, name and path fields")
281 parser.add_argument('--read1', action='store', dest='read1', required=False, default=None, help='Required: single read')
282 parser.add_argument('--read2', action='store', dest='read2', required=False, default=None, help='Optional: paired read')
283 parser.add_argument('--gzipped', action='store', dest='gzipped', help='Input files are gzipped')
284 parser.add_argument('--output_dbkey', action='store', dest='output_dbkey', required=False, default=None, help='Output reference file')
285 parser.add_argument('--output_metrics', action='store', dest='output_metrics', required=False, default=None, help='Output metrics file')
286 parser.add_argument('--processes', action='store', dest='processes', type=int, help='User-selected number of processes to use for job splitting')
287
288 args = parser.parse_args()
289
290 collection = False
291 fastq_list = []
292 if args.read1 is not None:
293 fastq_list.append(args.read1)
294 if args.read2 is not None:
295 fastq_list.append(args.read2)
296 else:
297 collection = True
298 for file_name in sorted(os.listdir(INPUT_READS_DIR)):
299 file_path = os.path.abspath(os.path.join(INPUT_READS_DIR, file_name))
300 fastq_list.append(file_path)
301
302 # The value of dnaprint_fields is a list of lists, where each list is
303 # the [value, name, path] components of the vsnp_dnaprints data table.
304 # The data_manager_vsnp_dnaprints tool assigns the dbkey column from the
305 # all_fasta data table to the value column in the vsnp_dnaprints data
306 # table to ensure a proper mapping for discovering the dbkey.
307 dnaprints_dict = get_dnaprints_dict(args.dnaprint_fields)
308
309 if collection:
310 # Here fastq_list consists of any number of
311 # reads, so each file will be processed and
312 # dataset collections will be produced as outputs.
313 multiprocessing.set_start_method('spawn')
314 queue1 = multiprocessing.JoinableQueue()
315 queue2 = multiprocessing.JoinableQueue()
316 num_files = len(fastq_list)
317 cpus = set_num_cpus(num_files, args.processes)
318 # Set a timeout for get()s in the queue.
319 timeout = 0.05
320
321 for fastq_file in fastq_list:
322 queue1.put(fastq_file)
323
324 # Complete the get_species_counts task.
325 processes = [multiprocessing.Process(target=get_species_counts_for_collection, args=(queue1, queue2, args.gzipped, timeout, )) for _ in range(cpus)]
326 for p in processes:
327 p.start()
328 for p in processes:
329 p.join()
330 queue1.join()
331
332 # Complete the get_species_strings task.
333 processes = [multiprocessing.Process(target=get_species_strings_for_collection, args=(queue2, queue1, timeout, )) for _ in range(cpus)]
334 for p in processes:
335 p.start()
336 for p in processes:
337 p.join()
338 queue2.join()
339
340 # Complete the get_group_and_dbkey task.
341 processes = [multiprocessing.Process(target=get_group_and_dbkey_for_collection, args=(queue1, queue2, dnaprints_dict, timeout, )) for _ in range(cpus)]
342 for p in processes:
343 p.start()
344 for p in processes:
345 p.join()
346 queue1.join()
347
348 # Complete the output_files task.
349 processes = [multiprocessing.Process(target=output_files_for_collection, args=(queue2, timeout, )) for _ in range(cpus)]
350 for p in processes:
351 p.start()
352 for p in processes:
353 p.join()
354 queue2.join()
355
356 if queue1.empty() and queue2.empty():
357 queue1.close()
358 queue1.join_thread()
359 queue2.close()
360 queue2.join_thread()
361 else:
362 # Here fastq_list consists of either a single read
363 # or a set of paired reads, producing single outputs.
364 count_summary, count_list, brucella_sum, bovis_sum, para_sum = get_species_counts(fastq_list, args.gzipped)
365 brucella_string, bovis_string, para_string = get_species_strings(count_summary)
366 group, dbkey = get_group_and_dbkey(dnaprints_dict, brucella_string, brucella_sum, bovis_string, bovis_sum, para_string, para_sum)
367 output_files(args.read1, count_list, group, dbkey, dbkey_file=args.output_dbkey, metrics_file=args.output_metrics)