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,
43 Util.check_file_for_readability( hmmsearch_output )
44 Util.check_file_for_readability( fasta_sequence_file )
45 Util.check_file_for_writability( outfile )
46 Util.check_file_for_writability( passed_seqs_outfile )
47 Util.check_file_for_writability( failed_seqs_outfile )
50 factory = MsaFactory.new()
51 in_msa = factory.create_msa_from_file( fasta_sequence_file, FastaParser.new() )
53 if ( in_msa == nil || in_msa.get_number_of_seqs() < 1 )
54 error_msg = "could not find fasta sequences in " + fasta_sequence_file
55 raise IOError, error_msg
62 ld = Constants::LINE_DELIMITER
64 domain_pass_counter = 0
65 domain_fail_counter = 0
66 proteins_with_passing_domains = 0
67 proteins_with_failing_domains = 0
68 max_domain_copy_number_per_protein = -1
69 max_domain_copy_number_sequence = ''
70 failed_species_counts = Hash.new
71 passed_species_counts = Hash.new
73 File.open( hmmsearch_output ) do | file |
74 while line = file.gets
75 if !is_ignorable?( line ) && line =~ /^\S+\s+/
77 # tn acc tlen query acc qlen Evalue score bias # of c-E i-E score bias hf ht af at ef et acc desc
78 # 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23
79 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+(.*)/
82 if domain_id != target_name
93 if ( number > max_domain_copy_number_per_protein )
94 max_domain_copy_number_sequence = sequence
95 max_domain_copy_number_per_protein = number
97 if ( ( ( e_value_threshold.to_f < 0.0 ) || ( i_e_value <= e_value_threshold ) ) &&
98 ( ( length_threshold.to_f <= 0 ) || ( env_to - env_from + 1 ) >= length_threshold.to_f ) )
99 extract_domain( sequence,
108 add_domain_number_as_digit,
109 add_domain_number_as_letter,
112 domain_pass_counter += 1
113 count_species( sequence, passed_species_counts )
114 if passed_seqs.find_by_name_start( sequence, true ).length < 1
115 add_sequence( sequence, in_msa, passed_seqs )
116 proteins_with_passing_domains += 1
119 print( domain_fail_counter.to_s + ": " + sequence.to_s + " did not meet threshold(s)" )
120 log << domain_fail_counter.to_s + ": " + sequence.to_s + " did not meet threshold(s)"
121 if ( ( e_value_threshold.to_f >= 0.0 ) && ( i_e_value > e_value_threshold ) )
122 print( " iE=" + i_e_value.to_s )
123 log << " iE=" + i_e_value.to_s
125 if ( ( length_threshold.to_f > 0 ) && ( env_to - env_from + 1 ) < length_threshold.to_f )
126 le = env_to - env_from + 1
127 print( " l=" + le.to_s )
128 log << " l=" + le.to_s
130 print( Constants::LINE_DELIMITER )
131 log << Constants::LINE_DELIMITER
132 domain_fail_counter += 1
133 count_species( sequence, failed_species_counts )
134 if failed_seqs.find_by_name_start( sequence, true ).length < 1
135 add_sequence( sequence, in_msa, failed_seqs )
136 proteins_with_failing_domains += 1
143 if domain_pass_counter < 1
144 error_msg = "no domain sequences were extracted"
145 raise StandardError, error_msg
148 log << Constants::LINE_DELIMITER
149 puts( "Max domain copy number per protein : " + max_domain_copy_number_per_protein.to_s )
150 log << "Max domain copy number per protein : " + max_domain_copy_number_per_protein.to_s
151 log << Constants::LINE_DELIMITER
153 if ( max_domain_copy_number_per_protein > 1 )
154 puts( "First protein with this copy number: " + max_domain_copy_number_sequence )
155 log << "First protein with this copy number: " + max_domain_copy_number_sequence
156 log << Constants::LINE_DELIMITER
160 w = FastaWriter.new()
161 w.set_line_width( 60 )
165 io.write_to_file( out_msa, outfile, w )
167 error_msg = "could not write to \"" + outfile + "\""
168 raise IOError, error_msg
172 io.write_to_file( passed_seqs, passed_seqs_outfile, w )
174 error_msg = "could not write to \"" + passed_seqs_outfile + "\""
175 raise IOError, error_msg
179 io.write_to_file( failed_seqs, failed_seqs_outfile, w )
181 error_msg = "could not write to \"" + failed_seqs_outfile + "\""
182 raise IOError, error_msg
186 log << "passing domains : " + domain_pass_counter.to_s + ld
187 log << "failing domains : " + domain_fail_counter.to_s + ld
188 log << "proteins with passing domains: " + proteins_with_passing_domains.to_s + ld
189 log << "proteins with failing domains: " + proteins_with_failing_domains.to_s + ld
191 log << 'passing domains counts per species: ' << ld
192 passed_species_counts.each_pair { | species, count | log << "#{species}: #{count}" << ld }
194 log << 'failing domains counts per species: ' << ld
195 failed_species_counts.each_pair { | species, count | log << "#{species}: #{count}" << ld }
197 return domain_pass_counter
204 def add_sequence( sequence_name, in_msa, add_to_msa )
205 seqs = in_msa.find_by_name_start( sequence_name, true )
206 if ( seqs.length < 1 )
207 error_msg = "sequence \"" + sequence_name + "\" not found in sequence file"
208 raise StandardError, error_msg
210 if ( seqs.length > 1 )
211 error_msg = "sequence \"" + sequence_name + "\" not unique in sequence file"
212 raise StandardError, error_msg
214 seq = in_msa.get_sequence( seqs[ 0 ] )
215 add_to_msa.add_sequence( seq )
218 # raises ArgumentError, StandardError
219 def extract_domain( sequence,
228 add_domain_number_as_digit,
229 add_domain_number_as_letter,
232 if ( number < 1 || out_of < 1 || number > out_of )
233 error_msg = "impossible: number=" + number.to_s + ", out of=" + out_of.to_s
234 raise ArgumentError, error_msg
236 if ( seq_from < 1 || seq_to < 1 || seq_from >= seq_to )
237 error_msg = "impossible: seq-f=" + seq_from.to_s + ", seq-t=" + seq_to.to_s
238 raise ArgumentError, error_msg
240 seqs = in_msa.find_by_name_start( sequence, true )
242 error_msg = "sequence \"" + sequence + "\" not found in sequence file"
243 raise StandardError, error_msg
246 error_msg = "sequence \"" + sequence + "\" not unique in sequence file"
247 raise StandardError, error_msg
249 # hmmsearch is 1 based, wheres sequences are 0 bases in this package.
250 seq = in_msa.get_sequence( seqs[ 0 ] ).get_subsequence( seq_from - 1, seq_to - 1 )
252 orig_name = seq.get_name
254 seq.set_name( orig_name.split[ 0 ] )
257 seq.set_name( seq.get_name + "_" + seq_from.to_s + "-" + seq_to.to_s )
261 seq.set_name( seq.get_name[ 0, seq.get_name.length - TRIM_BY ] )
265 if ( add_domain_number_as_digit )
266 seq.set_name( seq.get_name + number.to_s )
267 elsif ( add_domain_number_as_letter )
269 error_msg = 'too many identical domains per sequence, cannot use letters to distinguish them'
270 raise StandardError, error_msg
272 seq.set_name( seq.get_name + ( number + 96 ).chr )
273 elsif ( add_domain_number )
274 seq.set_name( seq.get_name + "~" + number.to_s + "-" + out_of.to_s )
278 a = orig_name.rindex "["
279 b = orig_name.rindex "]"
281 error_msg = "species not found in " + orig_name
282 raise StandardError, error_msg
284 species = orig_name[ a .. b ]
285 seq.set_name( seq.get_name + " [" + species + "]" )
290 # if ( seq.get_name.length > 10 )
291 # error_msg = "sequence name [" + seq.get_name + "] is longer than 10 characters"
292 # raise StandardError, error_msg
295 out_msa.add_sequence( seq )
298 def count_species( sequence, species_counts_map )
299 species = get_species( sequence )
301 if !species_counts_map.has_key?( species )
302 species_counts_map[ species ] = 1
304 species_counts_map[ species ] = species_counts_map[ species ] + 1
309 def get_species( sequence_name )
310 if sequence_name =~ /^.+_(.+)$/
317 def is_ignorable?( line )
318 return ( line !~ /[A-Za-z0-9-]/ || line =~/^#/ )
321 end # class HmmscanDomainExtractor