2 # = lib/evo/io/parser/hmmsearch_domain_extractor.rb - HmmsearchDomainExtractor class
4 # Copyright:: Copyright (C) 2006-2008 Christian M. Zmasek
5 # License:: GNU Lesser General Public License (LGPL)
7 # $Id: hmmsearch_domain_extractor.rb,v 1.24 2009/11/25 06:30:24 cmzmasek Exp $
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 HmmsearchDomainExtractor
26 # raises ArgumentError, IOError, StandardError
27 def parse( hmmsearch_output,
36 add_domain_number_as_digit,
37 add_domain_number_as_letter,
41 Util.check_file_for_readability( hmmsearch_output )
42 Util.check_file_for_readability( fasta_sequence_file )
43 Util.check_file_for_writability( outfile )
44 Util.check_file_for_writability( passed_seqs_outfile )
45 Util.check_file_for_writability( failed_seqs_outfile )
48 factory = MsaFactory.new()
49 in_msa = factory.create_msa_from_file( fasta_sequence_file, FastaParser.new() )
51 if ( in_msa == nil || in_msa.get_number_of_seqs() < 1 )
52 error_msg = "could not find fasta sequences in " + fasta_sequence_file
53 raise IOError, error_msg
60 ld = Constants::LINE_DELIMITER
62 domain_pass_counter = 0
63 domain_fail_counter = 0
64 proteins_with_passing_domains = 0
65 proteins_with_failing_domains = 0
66 max_domain_copy_number_per_protein = -1
67 max_domain_copy_number_sequence = ''
68 failed_species_counts = Hash.new
69 passed_species_counts = Hash.new
71 File.open( hmmsearch_output ) do | file |
72 while line = file.gets
73 if !is_ignorable?( line ) && line =~ /^\S+\s+/
75 # tn acc tlen query acc qlen Evalue score bias # of c-E i-E score bias hf ht af at ef et acc desc
76 # 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23
77 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+(.*)/
79 # line =~ /^(\S+)\s+(\d+)\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+)/
86 if ( number > max_domain_copy_number_per_protein )
87 max_domain_copy_number_sequence = sequence
88 max_domain_copy_number_per_protein = number
90 if ( ( ( e_value_threshold.to_f < 0.0 ) || ( i_e_value <= e_value_threshold ) ) &&
91 ( ( length_threshold.to_f <= 0 ) || ( env_to - env_from + 1 ) >= length_threshold.to_f ) )
92 HmmsearchDomainExtractor.extract_domain( sequence,
101 add_domain_number_as_digit,
102 add_domain_number_as_letter,
104 domain_pass_counter += 1
105 count_species( sequence, passed_species_counts )
106 if !passed_seqs.has?( sequence, true, false )
107 HmmsearchDomainExtractor.add_sequence( sequence, in_msa, passed_seqs )
108 proteins_with_passing_domains += 1
111 print( domain_fail_counter.to_s + ": " + sequence.to_s + " did not meet threshold(s)" )
112 log << domain_fail_counter.to_s + ": " + sequence.to_s + " did not meet threshold(s)"
113 if ( ( e_value_threshold.to_f >= 0.0 ) && ( i_e_value > e_value_threshold ) )
114 print( " iE=" + i_e_value.to_s )
115 log << " iE=" + i_e_value.to_s
117 if ( ( length_threshold.to_f > 0 ) && ( env_to - env_from + 1 ) < length_threshold.to_f )
118 le = env_to - env_from + 1
119 print( " l=" + le.to_s )
120 log << " l=" + le.to_s
122 print( Constants::LINE_DELIMITER )
123 log << Constants::LINE_DELIMITER
124 domain_fail_counter += 1
125 count_species( sequence, failed_species_counts )
126 if !failed_seqs.has?( sequence, true, false )
127 HmmsearchDomainExtractor.add_sequence( sequence, in_msa, failed_seqs )
128 proteins_with_failing_domains += 1
135 if domain_pass_counter < 1
136 error_msg = "no domain sequences were extracted"
137 raise StandardError, error_msg
140 log << Constants::LINE_DELIMITER
141 puts( "Max domain copy number per protein : " + max_domain_copy_number_per_protein.to_s )
142 log << "Max domain copy number per protein : " + max_domain_copy_number_per_protein.to_s
143 log << Constants::LINE_DELIMITER
145 if ( max_domain_copy_number_per_protein > 1 )
146 puts( "First protein with this copy number: " + max_domain_copy_number_sequence )
147 log << "First protein with this copy number: " + max_domain_copy_number_sequence
148 log << Constants::LINE_DELIMITER
152 w = FastaWriter.new()
153 w.set_line_width( 60 )
157 io.write_to_file( out_msa, outfile, w )
159 error_msg = "could not write to \"" + outfile + "\""
160 raise IOError, error_msg
164 io.write_to_file( passed_seqs, passed_seqs_outfile, w )
166 error_msg = "could not write to \"" + passed_seqs_outfile + "\""
167 raise IOError, error_msg
171 io.write_to_file( failed_seqs, failed_seqs_outfile, w )
173 error_msg = "could not write to \"" + failed_seqs_outfile + "\""
174 raise IOError, error_msg
178 log << "passing domains : " + domain_pass_counter.to_s + ld
179 log << "failing domains : " + domain_fail_counter.to_s + ld
180 log << "proteins with passing domains: " + proteins_with_passing_domains.to_s + ld
181 log << "proteins with failing domains: " + proteins_with_failing_domains.to_s + ld
183 log << 'passing domains counts per species: ' << ld
184 passed_species_counts.each_pair { | species, count | log << "#{species}: #{count}" << ld }
186 log << 'failing domains counts per species: ' << ld
187 failed_species_counts.each_pair { | species, count | log << "#{species}: #{count}" << ld }
189 return domain_pass_counter
196 def HmmsearchDomainExtractor.add_sequence( sequence_name, in_msa, add_to_msa )
197 seqs = in_msa.find_by_name( sequence_name, true, false )
198 if ( seqs.length < 1 )
199 error_msg = "sequence \"" + sequence_name + "\" not found in sequence file"
200 raise StandardError, error_msg
202 if ( seqs.length > 1 )
203 error_msg = "sequence \"" + sequence_name + "\" not unique in sequence file"
204 raise StandardError, error_msg
206 seq = in_msa.get_sequence( seqs[ 0 ] )
207 add_to_msa.add_sequence( seq )
210 # raises ArgumentError, StandardError
211 def HmmsearchDomainExtractor.extract_domain( sequence,
220 add_domain_number_as_digit,
221 add_domain_number_as_letter,
223 if ( number < 1 || out_of < 1 || number > out_of )
224 error_msg = "impossible: number=" + number.to_s + ", out of=" + out_of.to_s
225 raise ArgumentError, error_msg
227 if ( seq_from < 1 || seq_to < 1 || seq_from >= seq_to )
228 error_msg = "impossible: seq-f=" + seq_from.to_s + ", seq-t=" + seq_to.to_s
229 raise ArgumentError, error_msg
231 seqs = in_msa.find_by_name( sequence, true, false )
232 if ( seqs.length < 1 )
233 error_msg = "sequence \"" + sequence + "\" not found in sequence file"
234 raise StandardError, error_msg
236 if ( seqs.length > 1 )
237 error_msg = "sequence \"" + sequence + "\" not unique in sequence file"
238 raise StandardError, error_msg
240 # hmmsearch is 1 based, wheres sequences are 0 bases in this package.
241 seq = in_msa.get_sequence( seqs[ 0 ] ).get_subsequence( seq_from - 1, seq_to - 1 )
243 seq.set_name( seq.get_name + "_" + seq_from.to_s + "-" + seq_to.to_s )
247 seq.set_name( seq.get_name[ 0, seq.get_name.length - TRIM_BY ] )
251 if ( add_domain_number_as_digit )
252 seq.set_name( seq.get_name + number.to_s )
253 elsif ( add_domain_number_as_letter )
255 error_msg = 'too many identical domains per sequence, cannot use letters to distinguish them'
256 raise StandardError, error_msg
258 seq.set_name( seq.get_name + ( number + 96 ).chr )
259 elsif ( add_domain_number )
260 seq.set_name( seq.get_name + "~" + number.to_s + "-" + out_of.to_s )
264 if ( seq.get_name.length > 10 )
265 error_msg = "sequence name [" + seq.get_name + "] is longer than 10 characters"
266 raise StandardError, error_msg
269 out_msa.add_sequence( seq )
272 def count_species( sequence, species_counts_map )
273 species = get_species( sequence )
275 if !species_counts_map.has_key?( species )
276 species_counts_map[ species ] = 1
278 species_counts_map[ species ] = species_counts_map[ species ] + 1
283 def get_species( sequence_name )
284 if sequence_name =~ /^.+_(.+)$/
291 def is_ignorable?( line )
292 return ( line !~ /[A-Za-z0-9-]/ || line =~/^#/ )
295 end # class HmmsearchDomainExtractor