2 # = lib/evo/io/parser/hmmscan_domain_extractor.rb - HmmscanDomainExtractor class
4 # Copyright:: Copyright (C) 2012 Christian M. Zmasek
5 # License:: GNU Lesser General Public License (LGPL)
10 require 'lib/evo/util/constants'
11 require 'lib/evo/msa/msa_factory'
12 require 'lib/evo/io/msa_io'
13 require 'lib/evo/io/writer/fasta_writer'
14 require 'lib/evo/io/parser/fasta_parser'
19 class HmmscanDomainExtractor
26 # raises ArgumentError, IOError, StandardError
37 add_domain_number_as_digit,
38 add_domain_number_as_letter,
44 Util.check_file_for_readability( hmmsearch_output )
45 Util.check_file_for_readability( fasta_sequence_file )
46 Util.check_file_for_writability( outfile )
47 Util.check_file_for_writability( passed_seqs_outfile )
48 Util.check_file_for_writability( failed_seqs_outfile )
51 factory = MsaFactory.new()
52 in_msa = factory.create_msa_from_file( fasta_sequence_file, FastaParser.new() )
54 if ( in_msa == nil || in_msa.get_number_of_seqs() < 1 )
55 error_msg = "could not find fasta sequences in " + fasta_sequence_file
56 raise IOError, error_msg
64 out_msa_distance_partners = nil
65 out_msa_singlets = nil
67 out_msa_pairs = Msa.new
68 out_msa_distant_partners = Msa.new
69 out_msa_singlets = Msa.new
72 ld = Constants::LINE_DELIMITER
74 domain_pass_counter = 0
75 domain_fail_counter = 0
76 proteins_with_passing_domains = 0
77 proteins_with_failing_domains = 0
78 max_domain_copy_number_per_protein = -1
79 max_domain_copy_number_sequence = ""
88 File.open( hmmsearch_output ) do | file |
89 while line = file.gets
90 if !is_ignorable?( line ) && line =~ /^\S+\s+/
92 # tn acc tlen query acc qlen Evalue score bias # of c-E i-E score bias hf ht af at ef et acc desc
93 # 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23
94 line =~ /^(\S+)\s+(\S+)\s+(\d+)\s+(\S+)\s+(\S+)\s+(\d+)\s+(\S+)\s+(\S+)\s+(\S+)\s+(\d+)\s+(\d+)\s+(\S+)\s+(\S+)\s+(\S+)\s+(\S+)\s+(\d+)\s+(\d+)\s+(\d+)\s+(\d+)\s+(\d+)\s+(\d+)\s+(\S+)\s+(.*)/
97 if domain_id != target_name
107 if ( number > max_domain_copy_number_per_protein )
108 max_domain_copy_number_sequence = sequence
109 max_domain_copy_number_per_protein = number
111 if ( ( ( e_value_threshold < 0.0 ) || ( i_e_value <= e_value_threshold ) ) &&
112 ( ( length_threshold <= 0 ) || ( env_to - env_from + 1 ) >= length_threshold.to_f ) )
114 extract_domain( sequence,
123 add_domain_number_as_digit,
124 add_domain_number_as_letter,
127 domain_pass_counter += 1
129 if passed_seqs.find_by_name_start( sequence, true ).length < 1
130 add_sequence( sequence, in_msa, passed_seqs )
131 proteins_with_passing_domains += 1
135 if ( ( e_value_threshold < 0.0 ) || ( prev_i_e_value <= e_value_threshold ) ) &&
136 ( ( length_threshold <= 0 ) || ( ( prev_env_to - prev_env_from + 1 ) >= length_threshold.to_f ) )
138 if sequence != prev_sequence
144 if sequence == prev_sequence
145 puts "sequence == prev_sequence && out_of == 1"
148 extract_domain( sequence,
162 elsif sequence == prev_sequence
164 if ( env_from - prev_env_to ) <= min_linker #######
165 extract_domain( sequence,
166 prev_number.to_s + "+" + number.to_s,
181 extract_domain( sequence,
187 out_msa_distant_partners,
196 extract_domain( sequence,
202 out_msa_distant_partners,
214 prev_sequence = sequence
216 prev_env_from = env_from
218 prev_i_e_value = i_e_value
222 print( domain_fail_counter.to_s + ": " + sequence.to_s + " did not meet threshold(s)" )
223 log << domain_fail_counter.to_s + ": " + sequence.to_s + " did not meet threshold(s)"
224 if ( ( e_value_threshold.to_f >= 0.0 ) && ( i_e_value > e_value_threshold ) )
225 print( " iE=" + i_e_value.to_s )
226 log << " iE=" + i_e_value.to_s
228 if ( ( length_threshold.to_f > 0 ) && ( env_to - env_from + 1 ) < length_threshold.to_f )
229 le = env_to - env_from + 1
230 print( " l=" + le.to_s )
231 log << " l=" + le.to_s
233 print( Constants::LINE_DELIMITER )
234 log << Constants::LINE_DELIMITER
235 domain_fail_counter += 1
237 if failed_seqs.find_by_name_start( sequence, true ).length < 1
238 add_sequence( sequence, in_msa, failed_seqs )
239 proteins_with_failing_domains += 1
243 end # if !is_ignorable?( line ) && line =~ /^\S+\s+/
244 end # while line = file.gets
245 end # File.open( hmmsearch_output ) do | file |
247 if domain_pass_counter < 1
248 error_msg = "no domain sequences were extracted"
249 raise StandardError, error_msg
252 log << Constants::LINE_DELIMITER
253 puts( "Max domain copy number per protein : " + max_domain_copy_number_per_protein.to_s )
254 log << "Max domain copy number per protein : " + max_domain_copy_number_per_protein.to_s
255 log << Constants::LINE_DELIMITER
257 if ( max_domain_copy_number_per_protein > 1 )
258 puts( "First protein with this copy number: " + max_domain_copy_number_sequence )
259 log << "First protein with this copy number: " + max_domain_copy_number_sequence
260 log << Constants::LINE_DELIMITER
263 write_msa( out_msa, outfile )
264 write_msa( passed_seqs, passed_seqs_outfile )
265 write_msa( failed_seqs, failed_seqs_outfile )
268 write_msa( out_msa_pairs, outfile +"_" + min_linker.to_s )
272 write_msa( out_msa_singlets, outfile +"_singles" )
275 if out_msa_distant_partners
276 write_msa( out_msa_distant_partners, outfile +"_singles" )
281 log << "passing domains : " + domain_pass_counter.to_s + ld
282 log << "failing domains : " + domain_fail_counter.to_s + ld
283 log << "proteins with passing domains: " + proteins_with_passing_domains.to_s + ld
284 log << "proteins with failing domains: " + proteins_with_failing_domains.to_s + ld
287 return domain_pass_counter
294 def write_msa( msa, filename )
296 w = FastaWriter.new()
297 w.set_line_width( 60 )
300 io.write_to_file( msa, filename, w )
302 error_msg = "could not write to \"" + filename + "\""
303 raise IOError, error_msg
308 def add_sequence( sequence_name, in_msa, add_to_msa )
309 seqs = in_msa.find_by_name_start( sequence_name, true )
310 if ( seqs.length < 1 )
311 error_msg = "sequence \"" + sequence_name + "\" not found in sequence file"
312 raise StandardError, error_msg
314 if ( seqs.length > 1 )
315 error_msg = "sequence \"" + sequence_name + "\" not unique in sequence file"
316 raise StandardError, error_msg
318 seq = in_msa.get_sequence( seqs[ 0 ] )
319 add_to_msa.add_sequence( seq )
322 # raises ArgumentError, StandardError
323 def extract_domain( sequence,
332 add_domain_number_as_digit,
333 add_domain_number_as_letter,
336 if number.is_a?( Fixnum ) && ( number < 1 || out_of < 1 || number > out_of )
337 error_msg = "impossible: number=" + number.to_s + ", out of=" + out_of.to_s
338 raise ArgumentError, error_msg
340 if seq_from < 1 || seq_to < 1 || seq_from >= seq_to
341 error_msg = "impossible: seq-f=" + seq_from.to_s + ", seq-t=" + seq_to.to_s
342 raise ArgumentError, error_msg
344 seqs = in_msa.find_by_name_start( sequence, true )
346 error_msg = "sequence \"" + sequence + "\" not found in sequence file"
347 raise StandardError, error_msg
350 error_msg = "sequence \"" + sequence + "\" not unique in sequence file"
351 raise StandardError, error_msg
353 # hmmsearch is 1 based, wheres sequences are 0 bases in this package.
354 seq = in_msa.get_sequence( seqs[ 0 ] ).get_subsequence( seq_from - 1, seq_to - 1 )
356 orig_name = seq.get_name
358 seq.set_name( orig_name.split[ 0 ] )
361 seq.set_name( seq.get_name + "_" + seq_from.to_s + "-" + seq_to.to_s )
365 seq.set_name( seq.get_name[ 0, seq.get_name.length - TRIM_BY ] )
369 if add_domain_number_as_digit
370 seq.set_name( seq.get_name + number.to_s )
371 elsif add_domain_number_as_letter
373 error_msg = 'too many identical domains per sequence, cannot use letters to distinguish them'
374 raise StandardError, error_msg
376 seq.set_name( seq.get_name + ( number + 96 ).chr )
377 elsif add_domain_number
378 seq.set_name( seq.get_name + "~" + number.to_s + "-" + out_of.to_s )
382 # if ( seq.get_name.length > 10 )
383 # error_msg = "sequence name [" + seq.get_name + "] is longer than 10 characters"
384 # raise StandardError, error_msg
388 a = orig_name.rindex "["
389 b = orig_name.rindex "]"
391 error_msg = "species not found in " + orig_name
392 raise StandardError, error_msg
394 species = orig_name[ a .. b ]
395 seq.set_name( seq.get_name + " " + species )
397 out_msa.add_sequence( seq )
400 def is_ignorable?( line )
401 return ( line !~ /[A-Za-z0-9-]/ || line =~/^#/ )
404 end # class HmmscanDomainExtractor