2 # = lib/evo/tool/hmmscan_summary.rb - HmmscanSummary class
4 # Copyright:: Copyright (C) 2017 Christian M. Zmasek
5 # License:: GNU Lesser General Public License (LGPL)
7 require 'lib/evo/util/constants'
8 require 'lib/evo/util/util'
9 require 'lib/evo/util/command_line_arguments'
10 require 'lib/evo/io/parser/hmmscan_parser'
11 require 'lib/evo/msa/msa'
12 require 'lib/evo/msa/msa_factory'
13 require 'lib/evo/io/msa_io'
14 require 'lib/evo/io/parser/fasta_parser'
22 PRG_DESC = "hmmscan analysis"
24 COPYRIGHT = "2014 Christian Zmasek"
25 CONTACT = "phyloxml@gmail.com"
26 WWW = "https://sites.google.com/site/cmzmasek/home/software/forester"
28 I_E_VALUE_THRESHOLD_OPTION = "ie"
29 FS_E_VALUE_THRESHOLD_OPTION = "pe"
32 HELP_OPTION_1 = "help"
36 AVOID_HHMS = [ "RRM_1", "RRM_2", "RRM_3", "RRM_4", "RRM_5", "RRM_6" ]
37 LIMIT_FOR_CLOSE_DOMAINS = 20
42 Util.print_program_information( PRG_NAME,
52 cla = CommandLineArguments.new( ARGV )
53 rescue ArgumentError => e
54 Util.fatal_error( PRG_NAME, "error: " + e.to_s, STDOUT )
57 if ( cla.is_option_set?( HELP_OPTION_1 ) ||
58 cla.is_option_set?( HELP_OPTION_2 ) )
63 if ( cla.get_number_of_files != 2 && cla.get_number_of_files != 3 )
68 allowed_opts = Array.new
69 allowed_opts.push( I_E_VALUE_THRESHOLD_OPTION )
70 allowed_opts.push( FS_E_VALUE_THRESHOLD_OPTION )
71 allowed_opts.push( TARGET_MODELS )
72 allowed_opts.push( EXTRACTION )
74 disallowed = cla.validate_allowed_options_as_str( allowed_opts )
75 if ( disallowed.length > 0 )
76 Util.fatal_error( PRG_NAME,
77 "unknown option(s): " + disallowed,
81 inpath = cla.get_file_name( 0 )
82 Util.check_file_for_readability( inpath )
84 seq_file_path = cla.get_file_name( 1 )
85 Util.check_file_for_readability( seq_file_path )
87 extraction_output = nil
88 if ( cla.get_number_of_files == 3 )
89 extraction_output = cla.get_file_name( 2 )
90 if File.exist?( extraction_output )
91 Util.fatal_error( PRG_NAME, "error: [#{extraction_output}] already exists" )
97 if seq_file_path != nil
98 msa = read_fasta_file( seq_file_path )
101 i_e_value_threshold = -1
102 if ( cla.is_option_set?( I_E_VALUE_THRESHOLD_OPTION ) )
104 i_e_value_threshold = cla.get_option_value_as_float( I_E_VALUE_THRESHOLD_OPTION )
105 rescue ArgumentError => e
106 Util.fatal_error( PRG_NAME, "error: " + e.to_s, STDOUT )
108 if ( i_e_value_threshold < 0.0 )
109 Util.fatal_error( PRG_NAME, "attempt to use a negative i-E-value threshold", STDOUT )
113 fs_e_value_threshold = -1
114 if ( cla.is_option_set?( FS_E_VALUE_THRESHOLD_OPTION ) )
116 fs_e_value_threshold = cla.get_option_value_as_float( FS_E_VALUE_THRESHOLD_OPTION )
117 rescue ArgumentError => e
118 Util.fatal_error( PRG_NAME, "error: " + e.to_s, STDOUT )
120 if ( fs_e_value_threshold < 0.0 )
121 Util.fatal_error( PRG_NAME, "attempt to use a negative E-value threshold", STDOUT )
126 if ( cla.is_option_set?( TARGET_MODELS ) )
128 hmm_for_protein_output = cla.get_option_value( TARGET_MODELS )
129 target_models = hmm_for_protein_output.split( "/" )
130 rescue ArgumentError => e
131 Util.fatal_error( PRG_NAME, "error: " + e.to_s, STDOUT )
136 if ( cla.is_option_set?( EXTRACTION ) )
138 x_models = cla.get_option_value( EXTRACTION ).split( "/" )
139 rescue ArgumentError => e
140 Util.fatal_error( PRG_NAME, "error: " + e.to_s, STDOUT )
145 if i_e_value_threshold > 0
146 puts "i E-value threshold:" + "\t" + i_e_value_threshold.to_s
148 puts "i E-value threshold:" + "\t" + "n/a"
151 if fs_e_value_threshold > 0
152 puts "fs E-value threshold:" + "\t" + fs_e_value_threshold.to_s
154 puts "fs E-value threshold:" + "\t" + "n/a"
160 title << "SPECIES" + "\t"
161 target_models.each do | t |
164 title << "LENGTH" + "\t"
165 title << "#DOMAINS" + "\t"
166 title << "DOMAINS" + "\t"
167 unless x_models.empty?
168 title << "LINKERS" + "\t"
171 title << "DETAILED DA"
177 fs_e_value_threshold,
183 Util.fatal_error( PRG_NAME, "error: " + e.to_s, STDOUT )
191 def read_fasta_file( input )
195 msa = f.create_msa_from_file( input, FastaParser.new() )
196 rescue Exception => e
197 Util.fatal_error( PRG_NAME, "error: " + e.to_s )
203 # raises ArgumentError, IOError
206 fs_e_value_threshold,
212 extraction_output_file = nil
213 if extraction_output != nil
214 extraction_output_file = File.open( extraction_output, "a" )
217 hmmscan_parser = HmmscanParser.new( inpath )
219 results = hmmscan_parser.parse
221 hmmscan_results_per_protein = []
225 results.each do | r |
227 if !prev_query.empty? && prev_query != query
228 if !hmmscan_results_per_protein.empty?
229 process_hmmscan_results_per_protein( hmmscan_results_per_protein,
230 fs_e_value_threshold,
235 extraction_output_file )
237 hmmscan_results_per_protein.clear
242 if !AVOID_HHMS.include? r.model
243 hmmscan_results_per_protein << r
246 hmmscan_results_per_protein << r
251 if !hmmscan_results_per_protein.empty?
252 process_hmmscan_results_per_protein( hmmscan_results_per_protein,
253 fs_e_value_threshold,
258 extraction_output_file )
260 if extraction_output_file != nil
261 extraction_output_file.close
266 def process_hmmscan_results_per_protein( hmmscan_results_per_protein,
267 fs_e_value_threshold,
272 extraction_output_file )
274 raise StandardError, "target hmms is empty" if target_hmms.length < 1
275 raise StandardError, "results is empty" if hmmscan_results_per_protein.length < 1
276 raise StandardError, "msa is empty" if ( msa == nil || msa.get_number_of_seqs < 1 )
279 # filter according to i-Evalue threshold
280 # abort if fs Evalue too high
282 if fs_e_value_threshold >= 0
283 hmmscan_results_per_protein.each do | r |
284 target_hmms.each do | hmm |
285 if r.model == hmm && r.fs_e_value > fs_e_value_threshold
292 hmmscan_results_per_protein_filtered = []
295 hmmscan_results_per_protein.each do | r |
296 if i_e_value_threshold < 0 || r.i_e_value <= i_e_value_threshold
297 hmmscan_results_per_protein_filtered << r
298 target_hmms.each do | hmm |
307 if matched.length < target_hmms.length
310 if hmmscan_results_per_protein_filtered.length < 1
314 hmmscan_results_per_protein_filtered.sort! { |r1,r2| r1.env_from <=> r2.env_from }
317 target_hmms.each do | hmm |
318 hmmscan_results_per_protein_filtered.each do | r |
333 raise StandardError, "failed sanity check" if query != own.query || qlen != own.qlen
334 raise StandardError, "failed sanity check: qlen != own.qlen" if qlen != own.qlen
340 seq = get_sequence( query, msa )
341 species = get_species( seq )
342 raise StandardError, "could not get species" if species == nil || species.empty?
343 if x_models != nil && !x_models.empty?
344 ll = extract_linker( hmmscan_results_per_protein_filtered, x_models, seq, extraction_output_file )
350 s << own.fs_e_value.to_s + "\t"
353 s << qlen.to_s + "\t"
355 s << hmmscan_results_per_protein_filtered.length.to_s + "\t"
356 hmmscan_results_per_protein_filtered.each do | r |
367 s << make_overview_da( hmmscan_results_per_protein_filtered )
369 s << make_detailed_da( hmmscan_results_per_protein_filtered, qlen )
375 def extract_linker( hmmscan_results_per_protein_filtered, x_models, seq, extraction_output_file )
376 raise StandardError, "extraction output file is nil" if extraction_output_file == nil
378 hmmscan_results_per_protein_filtered.each do | r |
380 if ( x_models.length == 2 && prev_r.model == x_models[ 0 ] && r.model == x_models[ 1 ] )
381 ll = output_linker( prev_r.env_to, r.env_from, seq, extraction_output_file )
391 def get_sequence( query, msa )
393 indices = msa.find_by_name_start( query , true )
394 if indices.length != 1
395 if query[ -1, 1 ] == "|"
398 seq = msa.get_by_name_pattern( /\b#{Regexp.quote(query)}\b/ )
400 seq = msa.get_sequence( indices[ 0 ] )
406 def get_species( seq )
408 if seq.get_name =~ /\[([A-Z0-9]{3,5})\]/
415 def output_linker( first, last , seq, output_file )
416 if ( last - first >= 1 )
417 output_file.print( ">" + seq.get_name + " [" + first.to_s + "-" + last.to_s + "]" + "\n")
418 output_file.print( seq.get_subsequence( first - 1 , last - 1 ).get_sequence_as_string + "\n" )
425 def make_detailed_da( hmmscan_results_per_protein_filtered , qlen )
428 hmmscan_results_per_protein_filtered.each do | r |
430 s << make_interdomain_sequence( r.env_from - prev_r.env_to - 1 )
432 s << make_interdomain_sequence( r.env_from, false )
436 s << r.env_from.to_s << "-" << r.env_to.to_s
437 s << " " << r.i_e_value.to_s
441 s << make_interdomain_sequence( qlen - prev_r.env_to, false )
446 def make_overview_da( hmmscan_results_per_protein_filtered )
449 hmmscan_results_per_protein_filtered.each do | r |
453 if ( r.env_from - prev_r.env_to - 1 ) <= LIMIT_FOR_CLOSE_DOMAINS
454 overview << "~" << r.model
456 overview << "----" << r.model
466 def make_interdomain_sequence( d, mark_short = true )
485 puts " " + PRG_NAME + ".rb [options] <hmmscan outputfile> <sequences in fasta format> [outputfile]"
487 puts " options: -" + I_E_VALUE_THRESHOLD_OPTION + ": i-E-value threshold, default is no threshold"
488 puts " -" + FS_E_VALUE_THRESHOLD_OPTION + ": E-value threshold for full protein sequences, only for protein summary"
489 puts " -" + TARGET_MODELS + ": target HMMs (separated by /)"
490 puts " -" + EXTRACTION + ": to extract matching sequences to [outputfile]"