bacterial-annotator 0.3.2 → 0.3.3
Sign up to get free protection for your applications and to get access to all the features.
- checksums.yaml +4 -4
- data/lib/bacterial-comparator.rb +85 -7
- metadata +2 -2
checksums.yaml
CHANGED
@@ -1,7 +1,7 @@
|
|
1
1
|
---
|
2
2
|
SHA1:
|
3
|
-
metadata.gz:
|
4
|
-
data.tar.gz:
|
3
|
+
metadata.gz: 6a97539de7b1dff0da1ac8edaef6e0e10949c74e
|
4
|
+
data.tar.gz: 09fa573de375d54339adf3a8800318a49c059ba5
|
5
5
|
SHA512:
|
6
|
-
metadata.gz:
|
7
|
-
data.tar.gz:
|
6
|
+
metadata.gz: 608daa4501f95457e6d412fdc43d0672e74acd4ecf91186ea57da00614033e569688ca2d6a82f89f4266bc62da2ca20d7703a9ace0d04f131de57025c93a7685
|
7
|
+
data.tar.gz: 6ac5946ce49c23bfeb0d32dbc38c41516df960ab73af5f6a1a99689a08f7076ecffb2e361cf4f087d11dd1dcb74858ab35c84aaac687d9a8b28112eee765c05e
|
data/lib/bacterial-comparator.rb
CHANGED
@@ -23,6 +23,7 @@ class BacterialComparator
|
|
23
23
|
Dir.mkdir(@outdir) if ! Dir.exists? @outdir
|
24
24
|
@genomes_list = options[:genomes_list]
|
25
25
|
@proc = options[:proc].to_i
|
26
|
+
@phylo_nb_genes = options[:phylo_nb_genes]
|
26
27
|
|
27
28
|
min_cov = options[:min_cov].to_f
|
28
29
|
min_pid = options[:pidentity].to_f
|
@@ -199,25 +200,102 @@ class BacterialComparator
|
|
199
200
|
|
200
201
|
end
|
201
202
|
|
203
|
+
|
202
204
|
def mafft_align_all_pep
|
203
205
|
puts "# MAFFT multialign all protein sequences.."
|
204
206
|
Dir.chdir("#{@outdir}/genes-align-pep/")
|
205
|
-
|
206
|
-
|
207
|
-
|
207
|
+
|
208
|
+
is_done = 1
|
209
|
+
if Dir["*.pep"].length == Dir["*.aln"].length
|
210
|
+
Dir["*.aln"].each do |a|
|
211
|
+
if File.size(a) == 0
|
212
|
+
is_done = 0
|
213
|
+
end
|
214
|
+
end
|
215
|
+
else
|
216
|
+
is_done = 0
|
217
|
+
end
|
218
|
+
|
219
|
+
if is_done==0
|
220
|
+
Parallel.map(Dir["*.pep"], in_processes: @proc) { |f|
|
221
|
+
mafft_align f
|
222
|
+
}
|
223
|
+
end
|
224
|
+
|
225
|
+
concat_alignments "genes-align-pep.concat.fasta"
|
208
226
|
Dir.chdir("../../")
|
227
|
+
|
209
228
|
end
|
210
229
|
|
211
230
|
def mafft_align_all_dna
|
212
231
|
puts "# MAFFT multialign all gene sequences.."
|
213
|
-
puts "# MAFFT multialign all protein sequences.."
|
214
232
|
Dir.chdir("#{@outdir}/genes-align-dna/")
|
215
|
-
|
216
|
-
|
217
|
-
|
233
|
+
|
234
|
+
is_done = 1
|
235
|
+
if Dir["*.dna"].length == Dir["*.aln"].length
|
236
|
+
Dir["*.aln"].each do |a|
|
237
|
+
if File.size(a) == 0
|
238
|
+
is_done = 0
|
239
|
+
end
|
240
|
+
end
|
241
|
+
else
|
242
|
+
is_done = 0
|
243
|
+
end
|
244
|
+
|
245
|
+
if is_done == 0
|
246
|
+
Parallel.map(Dir["*.dna"], in_processes: @proc) { |f|
|
247
|
+
mafft_align f
|
248
|
+
}
|
249
|
+
end
|
250
|
+
|
251
|
+
concat_alignments "genes-align-dna.concat.fasta"
|
218
252
|
Dir.chdir("../../")
|
253
|
+
|
219
254
|
end
|
220
255
|
|
221
256
|
|
257
|
+
def concat_alignments outfile
|
258
|
+
|
259
|
+
fout = File.open("../#{outfile}", "w")
|
260
|
+
|
261
|
+
ref_id = Dir["../../#{@genomes_list[0]}/*.pep"][0].gsub(/.*\//,"").gsub(".pep","")
|
262
|
+
|
263
|
+
seq = ""
|
264
|
+
Dir["*.aln"].each do |f|
|
265
|
+
flat = Bio::FlatFile.auto(f)
|
266
|
+
ref_seq = flat.entries[0]
|
267
|
+
seq += ref_seq.seq
|
268
|
+
end
|
269
|
+
|
270
|
+
bioseq = Bio::Sequence.auto(seq)
|
271
|
+
out = bioseq.output_fasta("#{ref_id}",60)
|
272
|
+
fout.write(out)
|
273
|
+
|
274
|
+
for i in 1..@genomes_list.length
|
275
|
+
seq = ""
|
276
|
+
Dir["*.aln"].each do |f|
|
277
|
+
flat = Bio::FlatFile.auto(f)
|
278
|
+
j=0
|
279
|
+
flat.each_entry do |entry|
|
280
|
+
if j<i
|
281
|
+
j+=1
|
282
|
+
next
|
283
|
+
elsif i == j
|
284
|
+
seq += entry.seq
|
285
|
+
j+=1
|
286
|
+
else
|
287
|
+
break
|
288
|
+
end
|
289
|
+
end
|
290
|
+
end
|
291
|
+
bioseq = Bio::Sequence.auto(seq)
|
292
|
+
out = bioseq.output_fasta("#{@genomes_list[i-1]}",60)
|
293
|
+
fout.write(out)
|
294
|
+
end
|
295
|
+
|
296
|
+
fout.close
|
297
|
+
|
298
|
+
end
|
299
|
+
|
222
300
|
|
223
301
|
end # end of Class
|
metadata
CHANGED
@@ -1,14 +1,14 @@
|
|
1
1
|
--- !ruby/object:Gem::Specification
|
2
2
|
name: bacterial-annotator
|
3
3
|
version: !ruby/object:Gem::Version
|
4
|
-
version: 0.3.
|
4
|
+
version: 0.3.3
|
5
5
|
platform: ruby
|
6
6
|
authors:
|
7
7
|
- Maxime Deraspe
|
8
8
|
autorequire:
|
9
9
|
bindir: bin
|
10
10
|
cert_chain: []
|
11
|
-
date: 2016-10-
|
11
|
+
date: 2016-10-03 00:00:00.000000000 Z
|
12
12
|
dependencies:
|
13
13
|
- !ruby/object:Gem::Dependency
|
14
14
|
name: bio
|