|
0
|
1 '''
|
|
|
2 All configuration for clustering
|
|
|
3 '''
|
|
|
4 import os
|
|
|
5 import tempfile
|
|
|
6 from math import exp
|
|
|
7 from collections import namedtuple
|
|
|
8 MAIN_DIR = os.path.dirname(os.path.realpath(__file__))
|
|
|
9 def add_base_path(base):
|
|
|
10 '''automates generating absolute path in config'''
|
|
|
11 def joined_path(p):
|
|
|
12 '''create absolute path function '''
|
|
|
13 return os.path.join(base, p)
|
|
|
14 return joined_path
|
|
|
15
|
|
|
16 PATH = add_base_path(MAIN_DIR)
|
|
|
17
|
|
|
18 # clustering general settings
|
|
|
19 DIRECTORY_TREE = {'libdir': 'libdir',
|
|
|
20 'seqclust': 'seqclust',
|
|
|
21 'assembly': 'seqclust/small_clusters_assembly',
|
|
|
22 'blastx': 'seqclust/blastx',
|
|
|
23 'clustering': 'seqclust/clustering',
|
|
|
24 'clusters': 'seqclust/clustering/clusters',
|
|
|
25 'superclusters': 'seqclust/clustering/superclusters',
|
|
|
26 'mgblast': 'seqclust/mgblast',
|
|
|
27 'blastn': 'seqclust/blastn',
|
|
|
28 'prerun': 'seqclust/prerun',
|
|
|
29 'prerun_clusters': 'seqclust/prerun/clusters',
|
|
|
30 'sequences': 'seqclust/reads',
|
|
|
31 'custom_databases': 'seqclust/custom_databases'}
|
|
|
32
|
|
|
33 if "TEMP" in os.environ:
|
|
|
34 DIRECTORY_TREE['TEMP'] = os.environ["TEMP"]
|
|
|
35 else:
|
|
|
36 DIRECTORY_TREE['TEMP'] = tempfile.TemporaryDirectory().name
|
|
|
37
|
|
|
38 FILES = {'sample_db': DIRECTORY_TREE['TEMP'] + "/sample.db",
|
|
|
39 'sample_fasta': DIRECTORY_TREE['prerun'] + "/sample.fasta",
|
|
|
40 'prerun_cls_file' : DIRECTORY_TREE['prerun'] + "/sample_hitsort.cls",
|
|
|
41 'filter_sequences_file' : DIRECTORY_TREE['prerun'] + "/filter_sequences.fasta",
|
|
|
42 'sequences_db': DIRECTORY_TREE['TEMP'] + "/sequences.db",
|
|
|
43 'sequences_fasta': DIRECTORY_TREE['sequences'] + "/reads.fasta",
|
|
|
44 'hitsort': DIRECTORY_TREE['clustering'] + "/hitsort",
|
|
|
45 'hitsort_db': DIRECTORY_TREE['TEMP'] + "/hitsort.db",
|
|
|
46 'cls_file': DIRECTORY_TREE['clustering'] + "/hitsort.cls",
|
|
|
47 'clusters_summary_csv': "CLUSTER_TABLE.csv",
|
|
|
48 'profrep_classification_csv': "PROFREP_CLASSIFICATION_TEMPLATE.csv",
|
|
|
49 'superclusters_csv_summary': "SUPERCLUSTER_TABLE.csv",
|
|
|
50 'comparative_analysis_counts_csv': "COMPARATIVE_ANALYSIS_COUNTS.csv",
|
|
|
51 'clusters_info': ".clusters_info.csv",
|
|
|
52 'tarean_report_html': "tarean_report.html",
|
|
|
53 'cluster_report_html' : "cluster_report.html",
|
|
|
54 'supercluster_report_html' : 'supercluster_report.html',
|
|
|
55 'repeat_annotation_summary_rds' : 'repeat_annotation_summary.rds',
|
|
|
56 'summarized_annotation_html' :'summarized_annotation.html',
|
|
|
57 'main_report_html' : 'index.html',
|
|
|
58 'TR_consensus_fasta': "TAREAN_consensus_rank_{}.fasta",
|
|
|
59 'summary_histogram' : 'summary_histogram.png',
|
|
|
60 'comparative_summary_map': 'comparative_summary.png',
|
|
|
61 "how_to_cite" : "HOW_TO_CITE.html",
|
|
|
62 'logfile' : "logfile.txt",
|
|
|
63 'contigs' : "contigs.fasta",
|
|
|
64 'filter_omitted' : DIRECTORY_TREE['sequences'] + "/removed_filtering_positive_reads.fasta",
|
|
|
65 'filter_kept' : DIRECTORY_TREE['sequences'] + "/kept_filtering_positive_reads.fasta"
|
|
|
66 }
|
|
|
67
|
|
|
68
|
|
|
69 # include in output- [source, destination]
|
|
|
70 INCLUDE = [
|
|
|
71 [PATH("HOW_TO_CITE.html"), FILES["how_to_cite"]]
|
|
|
72 ]
|
|
|
73
|
|
|
74 # this is attribute of path - not a file name!
|
|
|
75 FILES_TO_DISCARD_AT_CLEANUP = [
|
|
|
76 'prerun', 'mgblast', 'blastn', "blastx",
|
|
|
77 'hitsort', "repeat_annotation_summary_rds"
|
|
|
78 ]
|
|
|
79
|
|
|
80 # relative links for html files
|
|
|
81 HTML_LINKS = {
|
|
|
82 "CLUSTER_TO_SUPERCLUSTER" : "../../superclusters/dir_SC%04d/index.html",
|
|
|
83 "SUPERCLUSTER_TO_CLUSTER" : "../../clusters/dir_CL%04d/index.html",
|
|
|
84 "CLUSTER_TO_CLUSTER" : "../dir_CL%04d/index.html",
|
|
|
85 "SUPERCLUSTER_TO_SUPERCLUSTER" : "../dir_SC%04d/index.html",
|
|
|
86 "CLUSTER_TO_CLUSTER_TABLE" : "../../../../cluster_report.html",
|
|
|
87 "SEPERCLUSTER_TO_CLUSTER_TABLE" : "../../../../cluster_report.html",
|
|
|
88 "ROOT_TO_CLUSTER" : "seqclust/clustering/clusters/dir_CL%04d/index.html",
|
|
|
89 "ROOT_TO_SUPERCLUSTER" : "seqclust/clustering/superclusters/dir_SC%04d/index.html",
|
|
|
90 "ROOT_TO_TAREAN" : "seqclust/clustering/clusters/dir_CL%04d/tarean/report.html",
|
|
|
91 "CLUSTER_TO_KMER_REPORT" : "tarean/report.html",
|
|
|
92 "INDEX_TO_TAREAN": "tarean_report.html",
|
|
|
93 "INDEX_TO_CLUSTER_REPORT": "cluster_report.html",
|
|
|
94 "INDEX_TO_SUPERCLUSTER_REPORT" : "supercluster_report.html",
|
|
|
95 "INDEX_TO_SUMMARIZED_ANNOTATION" : "summarized_annotation.html"
|
|
|
96 }
|
|
|
97
|
|
|
98
|
|
|
99 EMAX = 42.6 # define how many graph edges can be processed in 1Kb RAM
|
|
|
100 # MINIMUM_NUMBER_OF_INPUT_SEQUENCES = 5000
|
|
|
101 # FOR TESTING:
|
|
|
102 MINIMUM_NUMBER_OF_INPUT_SEQUENCES = 1000
|
|
|
103 MINIMUM_NUMBER_OF_READS_IN_CLUSTER = 20 # smaller clusters are not analyzed
|
|
|
104 MINIMUM_NUMBER_OF_READS_FOR_MERGING = 20 # smaller clusters will not be merged
|
|
|
105 MINIMUM_NUMBER_OF_SHARED_PAIRS_FOR_MERGING = 20 # min size of W param
|
|
|
106
|
|
|
107
|
|
|
108 NUMBER_OF_SEQUENCES_FOR_PRERUN_WITH_FILTERING = 50000
|
|
|
109 NUMBER_OF_SEQUENCES_FOR_PRERUN_WITHOUT_FILTERING = 20000
|
|
|
110 NUMBER_OF_SEQUENCES_FOR_PRERUN = NUMBER_OF_SEQUENCES_FOR_PRERUN_WITHOUT_FILTERING
|
|
|
111 CHUNK_SIZE = 20000
|
|
|
112
|
|
|
113
|
|
|
114 CLUSTER_EMAX = 2E7 # this parameter higle affect memory usage!
|
|
|
115
|
|
|
116 CLUSTER_VMAX = 40000
|
|
|
117 SUPERCLUSTER_THRESHOLD = 0.1
|
|
|
118 # Number of processors to use - it will be set at runtime
|
|
|
119 PROC = None
|
|
|
120 RSERVE_PORT = 6311
|
|
|
121
|
|
|
122 #some settings related to repeats annotation
|
|
|
123 ORF_THRESHOLD = 1200
|
|
|
124 PBS_THRESHOLD = 2
|
|
|
125 # threshold for rDNA detection - percentage of similarity hits
|
|
|
126 RDNA_THRESHOLD = 20
|
|
|
127 # threshold for contamination detection
|
|
|
128 CONTAMINATION_THRESHOLD = 10
|
|
|
129 # tandem ranks codes:
|
|
|
130 # 1 : putative tandem repeats - high confidence
|
|
|
131 # 2 : putative tandem repeats - low confidence
|
|
|
132 # 3 : potential LTR element
|
|
|
133 # 4 : rDNA
|
|
|
134 TANDEM_RANKS = [1, 2, 3, 4]
|
|
|
135 SKIP_CAP3_ASSEMBLY_TANDEM_RANKS = [1]
|
|
|
136 FILTER_MIN_PROP_THRESHOLD = 0.03 # this is minimal proportion of graph edges!
|
|
|
137 FILTER_MIN_SIZE_THRESHOLD = 1000 # minimal size of the cluster to be consider for filtering
|
|
|
138 FILTER_PROPORTION_OF_KEPT = 0.1
|
|
|
139
|
|
|
140 R = 'lib'
|
|
|
141 # external scripts
|
|
|
142 RSOURCE_tarean = PATH('lib/tarean/tarean.R')
|
|
|
143 RSOURCE_reporting = PATH('lib/reporting.R')
|
|
|
144 RSOURCE_create_annotation = PATH('lib/create_annotation.R')
|
|
|
145 LTR_DETECTION = PATH("lib/detect_LTR_insertion_sites.pl")
|
|
|
146
|
|
|
147 #PATH to DATABASES:
|
|
|
148 DNA_DATABASE = PATH("databases/dna_database_masked.fasta")
|
|
|
149 TRNA_DATABASE = PATH("databases/tRNA_database.fasta")
|
|
|
150 SATELLITE_MODEL = PATH("databases/satellite_model.rds")
|
|
|
151 LASTAL_PARAMS = PATH("databases/lastal_params")
|
|
|
152 # for testing
|
|
|
153 PROTEIN_DATABASE = None
|
|
|
154 CLASSIFICATION_HIERARCHY = None
|
|
|
155 CUSTOM_DNA_DATABASE = None
|
|
|
156
|
|
|
157 # when modifying this section check if makefile has most recent target for protein database
|
|
|
158 PROTEIN_DATABASE_DEFAULT = "VIRIDIPLANTAE3.0"
|
|
|
159 PROTEIN_DATABASE_OPTIONS = {'VIRIDIPLANTAE3.0' :
|
|
|
160 (PATH("databases/protein_database_viridiplantae_v3.0.fasta"), # change according if you use custom protein database
|
|
|
161 PATH("databases/classification_tree_viridiplantae_v3.0.rds")), # classification schem - data.tree object
|
|
|
162 'VIRIDIPLANTAE2.2' :
|
|
|
163 (PATH("databases/protein_database_viridiplantae_v2.2.fasta"), # change according if you use custom protein database
|
|
|
164 PATH("databases/classification_viridiplantae_tree.rds")), # classification schem - data.tree object
|
|
|
165 'METAZOA2.0' :
|
|
|
166 (PATH("databases/protein_database_metazoa_v3.fasta"), # change according if you use custom protein database
|
|
|
167 PATH("databases/classification_tree_metazoa_v3.rds")), # classification schem - data.tree object
|
|
|
168 'METAZOA3.0' :
|
|
|
169 (PATH("databases/protein_database_metazoa_v3.fasta"), # change according if you use custom protein database
|
|
|
170 PATH("databases/classification_tree_metazoa_v3.rds")) # classification schem - data.tree object
|
|
|
171 }
|
|
|
172 # if you change PROTEIN_DATABASE_OPTIONS, do not forget to use 'makeblastdb' build blast database
|
|
|
173 # and 'diamond makedb' to build diamond database
|
|
|
174
|
|
|
175 # PATH to binaries
|
|
|
176 LOUVAIN = PATH("louvain")
|
|
|
177 BINARIES = PATH("bin")
|
|
|
178
|
|
|
179 CAP3_PATTERNS_REPLACE = {"{}.{}.contigs" : [">Contig", ">{}Contig"],
|
|
|
180 "{}.{}.aln" : ["* Contig", "* {}Contig"],
|
|
|
181 "{}.{}.contigs.qual" : [">Contig", ">{}Contig"],
|
|
|
182 "{}.{}.ace" : ["CO Contig", "CO {}Contig"],
|
|
|
183 "{}.{}.singlets" : None,
|
|
|
184 "{}.{}.info" : None,
|
|
|
185 "{}.{}.contigs.links" : None}
|
|
|
186
|
|
|
187 CAP3_FILES_MAPPING = {"{}.{}.contigs" : "small_clusters.fasta",
|
|
|
188 "{}.{}.contigs.qual" : "small_clusters.contigs.qual",
|
|
|
189 "{}.{}.aln" : "small_clusters.aln",
|
|
|
190 "{}.{}.ace" : "small_clusters.ace",
|
|
|
191 "{}.{}.singlets" : None,
|
|
|
192 "{}.{}.info" : None,
|
|
|
193 "{}.{}.contigs.links" : None}
|
|
|
194
|
|
|
195 CAP3_FILENAMES = list(CAP3_PATTERNS_REPLACE.keys())
|
|
|
196
|
|
|
197 CAP3_FILES_GOODNAMES = {"{}.{}.contigs" : "contigs.fasta",
|
|
|
198 "{}.{}.contigs.qual" : "contigs.qual",
|
|
|
199 "{}.{}.aln" : "contigs.aln",
|
|
|
200 "{}.{}.ace" : "contigs.ace",
|
|
|
201 "{}.{}.singlets" : "singlets.fasta",
|
|
|
202 "{}.{}.info" : "assembly.info",
|
|
|
203 "{}.{}.contigs.links" : "contigs.links"}
|
|
|
204
|
|
|
205
|
|
|
206
|
|
|
207
|
|
|
208 CAP3_PARAMS = " -p 80 -o 40 " ## not implemented yet
|
|
|
209
|
|
|
210 LTR_DETECTION_FILES = {'ADJ': 'LTR_info.ADJ',
|
|
|
211 'LTR': 'LTR_info.LTR',
|
|
|
212 'PBS_BLAST': 'LTR_info.with_PBS_blast.csv',
|
|
|
213 'BASE' : 'LTR_info'}
|
|
|
214
|
|
|
215 # options for all-2-all search and annotations
|
|
|
216 FilteringThreshod = namedtuple("FilteringThreshold",
|
|
|
217 "min_lcov min_pid min_ovl min_scov evalue")
|
|
|
218 AnnotationParams = namedtuple("AnnotationParams", "blastn blastx blastn_trna")
|
|
|
219 Option = namedtuple('Options',
|
|
|
220 ('name database all2all_search_params filtering_threshold '
|
|
|
221 'filter_self_hits legacy_database lastdb annotation_search_params'))
|
|
|
222
|
|
|
223 # protein domain search options:
|
|
|
224 DIAMOND = {
|
|
|
225 'args': ' -p {max_proc} --max-target-seqs 1 --min-score 30 --freq-sd 1000 --more-sensitive',
|
|
|
226 'output_columns' : "qseqid sseqid qlen slen length ppos bitscore",
|
|
|
227 'column_types' : [str, str, float, float, float, float, float],
|
|
|
228 'program': 'diamond blastx',
|
|
|
229 'filter_function' : lambda x: x.bitscore >= 30,
|
|
|
230 'parallelize' : False
|
|
|
231 }
|
|
|
232 BLASTX_W3 = {
|
|
|
233 'args': ' -num_alignments 1 -word_size 2 -evalue 0.01 ',
|
|
|
234 'output_columns' : "qseqid sseqid qlen slen length ppos bitscore",
|
|
|
235 'column_types' : [str, str, float, float, float, float, float],
|
|
|
236 'program': 'blastx',
|
|
|
237 'filter_function' : lambda x: x.bitscore >= 33
|
|
|
238 }
|
|
|
239 BLASTX_W2 = BLASTX_W3
|
|
|
240 BLASTX_W2['args'] = ' -num_alignments 1 -word_size 3 -evalue 0.01 '
|
|
|
241
|
|
|
242
|
|
|
243 ARGS = None
|
|
|
244
|
|
|
245 ILLUMINA = Option(
|
|
|
246 name="illumina",
|
|
|
247 database='blastdb_legacy',
|
|
|
248 all2all_search_params=('mgblast -p 75 -W18 -UT -X40 -KT -JF -F '
|
|
|
249 '"m D" -v100000000 -b100000000'
|
|
|
250 ' -D4 -C 30 -H 30 -i {query} -d {blastdb}'),
|
|
|
251 filtering_threshold=FilteringThreshod(55, 90, 0, 0, 1),
|
|
|
252 filter_self_hits=False,
|
|
|
253 legacy_database=True,
|
|
|
254 lastdb=False,
|
|
|
255 annotation_search_params=AnnotationParams(
|
|
|
256 blastn={
|
|
|
257 'args': ' -task blastn -num_alignments 1 -evalue 0.01 ',
|
|
|
258 'output_columns' : "qseqid sseqid qlen slen length ppos bitscore",
|
|
|
259 'column_types' : [str, str, float, float, float, float, float],
|
|
|
260 'program': 'blastn',
|
|
|
261 'filter_function' : lambda x: x.length > 30 and x.bitscore > 60
|
|
|
262
|
|
|
263 },
|
|
|
264 blastx=BLASTX_W3,
|
|
|
265 blastn_trna={
|
|
|
266 'args': ' -task blastn -num_alignments 1 -word_size 7',
|
|
|
267 'output_columns' : "qseqid sseqid qlen slen length ppos bitscore",
|
|
|
268 'column_types' : [str, str, float, float, float, float, float],
|
|
|
269 'program': 'blastn',
|
|
|
270 'filter_function' : lambda x: x.length > 18 and x.bitscore > 60
|
|
|
271 }
|
|
|
272 )
|
|
|
273 )
|
|
|
274
|
|
|
275 ILLUMINA_DUST_OFF = ILLUMINA._replace(
|
|
|
276 all2all_search_params=('mgblast -p 75 -W18 -UT -X40 -KT -JF -F '
|
|
|
277 'F -v100000000 -b100000000'
|
|
|
278 ' -D4 -C 30 -H 30 -i {query} -d {blastdb}'),
|
|
|
279 )
|
|
|
280
|
|
|
281 ILLUMINA_SHORT = ILLUMINA._replace(
|
|
|
282 name="illumina_short",
|
|
|
283 all2all_search_params=('mgblast -p 75 -W18 -UT -X40 -KT -JF -F '
|
|
|
284 '"m D" -v100000000 -b100000000'
|
|
|
285 ' -D4 -C 20 -H 30 -i {query} -d {blastdb}'),
|
|
|
286 filtering_threshold=FilteringThreshod(40, 90, 0, 0, 0.1)
|
|
|
287 )
|
|
|
288
|
|
|
289
|
|
|
290 OXFORD_NANOPORE = Option(
|
|
|
291 name="oxford_nanopore",
|
|
|
292 database='lastdb',
|
|
|
293 all2all_search_params=('last_wrapper.py -f blasttab+ -P1 '
|
|
|
294 ' -m 700 -p {} '
|
|
|
295 ' {{blastdb}} {{query}} ').format(LASTAL_PARAMS),
|
|
|
296 filtering_threshold=FilteringThreshod(40, 50, 0, 0, 0.01),
|
|
|
297 filter_self_hits=True,
|
|
|
298 legacy_database=False,
|
|
|
299 lastdb=True,
|
|
|
300 annotation_search_params=AnnotationParams(
|
|
|
301 blastn={
|
|
|
302 'args': ' -task blastn -num_alignments 1 -evalue 0.01 -word_size 11',
|
|
|
303 'output_columns' : "qseqid sseqid qlen slen length ppos bitscore",
|
|
|
304 'column_types' : [str, str, float, float, float, float, float],
|
|
|
305 'program': 'blastn',
|
|
|
306 'filter_function' : lambda x: x.length > 30 and x.bitscore > 50
|
|
|
307 },
|
|
|
308 blastx={
|
|
|
309 'args': ' -num_alignments 1 -word_size 2 -evalue 0.1',
|
|
|
310 'output_columns' : "qseqid sseqid qlen slen length ppos bitscore",
|
|
|
311 'column_types' : [str, str, float, float, float, float, float],
|
|
|
312 'program': 'blastx',
|
|
|
313 'filter_function' : lambda x: x.bitscore >= 30
|
|
|
314 },
|
|
|
315 blastn_trna={
|
|
|
316 'args': ' -task blastn -num_alignments 1 -word_size 7',
|
|
|
317 'output_columns' : "qseqid sseqid qlen slen length ppos bitscore",
|
|
|
318 'column_types' : [str, str, float, float, float, float, float],
|
|
|
319 'program': 'blastn',
|
|
|
320 'filter_function' : lambda x: x.length > 18 and x.length > 60
|
|
|
321 }
|
|
|
322 )
|
|
|
323 )
|
