class HmmscanDomainExtractor
- TRIM_BY = 2
-
def initialize
end
# raises ArgumentError, IOError, StandardError
def parse( domain_id,
- hmmsearch_output,
+ hmmscan_output,
fasta_sequence_file,
outfile,
passed_seqs_outfile,
length_threshold,
add_position,
add_domain_number,
- add_domain_number_as_digit,
- add_domain_number_as_letter,
- trim_name,
add_species,
min_linker,
log )
- Util.check_file_for_readability( hmmsearch_output )
+ Util.check_file_for_readability( hmmscan_output )
Util.check_file_for_readability( fasta_sequence_file )
- Util.check_file_for_writability( outfile )
+ Util.check_file_for_writability( outfile + ".fasta" )
Util.check_file_for_writability( passed_seqs_outfile )
Util.check_file_for_writability( failed_seqs_outfile )
failed_seqs = Msa.new
passed_seqs = Msa.new
out_msa_pairs = nil
- out_msa_distance_partners = nil
- out_msa_singlets = nil
+ out_msa_isolated = nil
+ out_msa_singles = nil
if min_linker
out_msa_pairs = Msa.new
- out_msa_distant_partners = Msa.new
- out_msa_singlets = Msa.new
+ out_msa_isolated = Msa.new
+ out_msa_singles = Msa.new
end
ld = Constants::LINE_DELIMITER
domain_pass_counter = 0
domain_fail_counter = 0
- proteins_with_passing_domains = 0
proteins_with_failing_domains = 0
max_domain_copy_number_per_protein = -1
max_domain_copy_number_sequence = ""
- prev_sequence = ""
- prev_number = nil
- prev_env_from = nil
- prev_env_to = nil
- prev_i_e_value = nil
- prev_is_pair = false
+ hmmscan_datas = Array.new
- File.open( hmmsearch_output ) do | file |
+ File.open( hmmscan_output ) do | file |
while line = file.gets
if !is_ignorable?( line ) && line =~ /^\S+\s+/
# 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23
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+(.*)/
- target_name = $1
- if domain_id != target_name
+ if domain_id != $1
next
end
env_from = $20.to_i
env_to = $21.to_i
i_e_value = $13.to_f
- if ( number > max_domain_copy_number_per_protein )
- max_domain_copy_number_sequence = sequence
- max_domain_copy_number_per_protein = number
- end
- if ( ( ( e_value_threshold < 0.0 ) || ( i_e_value <= e_value_threshold ) ) &&
- ( ( length_threshold <= 0 ) || ( env_to - env_from + 1 ) >= length_threshold.to_f ) )
-
- extract_domain( sequence,
- number,
- out_of,
- env_from,
- env_to,
- in_msa,
- out_msa,
- add_position,
- add_domain_number,
- add_domain_number_as_digit,
- add_domain_number_as_letter,
- trim_name ,
- add_species )
- domain_pass_counter += 1
- if passed_seqs.find_by_name_start( sequence, true ).length < 1
- add_sequence( sequence, in_msa, passed_seqs )
- proteins_with_passing_domains += 1
+ if ( ( ( e_value_threshold < 0.0 ) || ( i_e_value <= e_value_threshold ) ) &&
+ ( ( length_threshold <= 0 ) || ( env_to - env_from + 1 ) >= length_threshold.to_f ) )
+ hmmscan_datas << HmmsearchData.new( sequence, number, out_of, env_from, env_to, i_e_value )
+ if ( number > max_domain_copy_number_per_protein )
+ max_domain_copy_number_sequence = sequence
+ max_domain_copy_number_per_protein = number
+ end
+ else # failed
+ print( domain_fail_counter.to_s + ": " + sequence.to_s + " did not meet threshold(s)" )
+ log << domain_fail_counter.to_s + ": " + sequence.to_s + " did not meet threshold(s)"
+ if ( ( e_value_threshold.to_f >= 0.0 ) && ( i_e_value > e_value_threshold ) )
+ print( " iE=" + i_e_value.to_s )
+ log << " iE=" + i_e_value.to_s
end
+ if ( ( length_threshold.to_f > 0 ) && ( env_to - env_from + 1 ) < length_threshold.to_f )
+ le = env_to - env_from + 1
+ print( " l=" + le.to_s )
+ log << " l=" + le.to_s
+ end
+ print( ld )
+ log << ld
+ domain_fail_counter += 1
+ end
- if min_linker
- if ( ( e_value_threshold < 0.0 ) || ( prev_i_e_value <= e_value_threshold ) ) &&
- ( ( length_threshold <= 0 ) || ( ( prev_env_to - prev_env_from + 1 ) >= length_threshold.to_f ) )
-
- if sequence != prev_sequence
- prev_is_pair = false
- end
-
- if out_of == 1
-
- if sequence == prev_sequence
- puts "sequence == prev_sequence && out_of == 1"
- exit
- end
- extract_domain( sequence,
- number,
- out_of,
- env_from,
- env_to,
- in_msa,
- out_msa_singlets,
- false,
- true,
- false,
- false,
- trim_name ,
- add_species )
-
- elsif sequence == prev_sequence
-
- if ( env_from - prev_env_to ) <= min_linker #######
- extract_domain( sequence,
- prev_number.to_s + "+" + number.to_s,
- out_of,
- prev_env_from,
- env_to,
- in_msa,
- out_msa_pairs,
- false,
- true,
- false,
- false,
- trim_name ,
- add_species )
- prev_is_pair = true
- else #######
- if !prev_is_pair
- extract_domain( sequence,
- prev_number,
- out_of,
- prev_env_from,
- prev_env_to,
- in_msa,
- out_msa_distant_partners,
- false,
- true,
- false,
- false,
- trim_name ,
- add_species )
- end
- if number == out_of
- extract_domain( sequence,
- number,
- out_of,
- env_from,
- env_to,
- in_msa,
- out_msa_distant_partners,
- false,
- true,
- false,
- false,
- trim_name ,
- add_species )
- end
- prev_is_pair = false
- end #######
-
- end
- prev_sequence = sequence
- prev_number = number
- prev_env_from = env_from
- prev_env_to = env_to
- prev_i_e_value = i_e_value
- end
+ if number > out_of
+ error_msg = "number > out_of ! (this should not have happened)"
+ raise StandardError, error_msg
+ end
- else
- print( domain_fail_counter.to_s + ": " + sequence.to_s + " did not meet threshold(s)" )
- log << domain_fail_counter.to_s + ": " + sequence.to_s + " did not meet threshold(s)"
- if ( ( e_value_threshold.to_f >= 0.0 ) && ( i_e_value > e_value_threshold ) )
- print( " iE=" + i_e_value.to_s )
- log << " iE=" + i_e_value.to_s
- end
- if ( ( length_threshold.to_f > 0 ) && ( env_to - env_from + 1 ) < length_threshold.to_f )
- le = env_to - env_from + 1
- print( " l=" + le.to_s )
- log << " l=" + le.to_s
+ if number == out_of
+ if !hmmscan_datas.empty?
+ process_hmmscan_datas( hmmscan_datas,
+ in_msa,
+ add_position,
+ add_domain_number,
+ add_species,
+ out_msa,
+ out_msa_singles,
+ out_msa_pairs,
+ out_msa_isolated,
+ min_linker )
+ domain_pass_counter += hmmscan_datas.length
+ if passed_seqs.find_by_name_start( sequence, true ).length < 1
+ add_sequence( sequence, in_msa, passed_seqs )
+ else
+ error_msg = "this should not have happened"
+ raise StandardError, error_msg
end
- print( Constants::LINE_DELIMITER )
- log << Constants::LINE_DELIMITER
- domain_fail_counter += 1
-
+ else
if failed_seqs.find_by_name_start( sequence, true ).length < 1
add_sequence( sequence, in_msa, failed_seqs )
proteins_with_failing_domains += 1
+ else
+ error_msg = "this should not have happened"
+ raise StandardError, error_msg
end
end
+ hmmscan_datas.clear
end
- end # if !is_ignorable?( line ) && line =~ /^\S+\s+/
+
+ end
end # while line = file.gets
- end # File.open( hmmsearch_output ) do | file |
+ end # File.open( hmmsearch_output ) do | file |
+
+
if domain_pass_counter < 1
error_msg = "no domain sequences were extracted"
- raise StandardError, error_msg
+ raise IOError, error_msg
end
- log << Constants::LINE_DELIMITER
+ log << ld
puts( "Max domain copy number per protein : " + max_domain_copy_number_per_protein.to_s )
log << "Max domain copy number per protein : " + max_domain_copy_number_per_protein.to_s
- log << Constants::LINE_DELIMITER
+ log << ld
if ( max_domain_copy_number_per_protein > 1 )
puts( "First protein with this copy number: " + max_domain_copy_number_sequence )
log << "First protein with this copy number: " + max_domain_copy_number_sequence
- log << Constants::LINE_DELIMITER
+ log << ld
end
- write_msa( out_msa, outfile )
+ write_msa( out_msa, outfile + ".fasta" )
write_msa( passed_seqs, passed_seqs_outfile )
write_msa( failed_seqs, failed_seqs_outfile )
if out_msa_pairs
- write_msa( out_msa_pairs, outfile +"_" + min_linker.to_s )
+ write_msa( out_msa_pairs, outfile +"_" + min_linker.to_s + ".fasta")
end
- if out_msa_singlets
- write_msa( out_msa_singlets, outfile +"_singles" )
+ if out_msa_singles
+ write_msa( out_msa_singles, outfile +"_singles" + ".fasta")
end
- if out_msa_distant_partners
- write_msa( out_msa_distant_partners, outfile +"_singles" )
+ if out_msa_isolated
+ write_msa( out_msa_isolated, outfile + "_" + min_linker.to_s + "_isolated" + ".fasta");
end
log << ld
- log << "passing domains : " + domain_pass_counter.to_s + ld
- log << "failing domains : " + domain_fail_counter.to_s + ld
- log << "proteins with passing domains: " + proteins_with_passing_domains.to_s + ld
- log << "proteins with failing domains: " + proteins_with_failing_domains.to_s + ld
+ log << "passing domains : " + domain_pass_counter.to_s + ld
+ if ( min_linker )
+ log << "single domains : " + out_msa_singles.get_number_of_seqs.to_s + ld
+ log << "domains in close pairs : " + out_msa_pairs.get_number_of_seqs.to_s + ld
+ log << "isolated domains : " + out_msa_isolated.get_number_of_seqs.to_s + ld
+ end
+ log << "failing domains : " + domain_fail_counter.to_s + ld
+ log << "proteins with passing domains : " + passed_seqs.get_number_of_seqs.to_s + ld
+ log << "proteins with no passing domains: " + proteins_with_failing_domains.to_s + ld
log << ld
return domain_pass_counter
add_to_msa.add_sequence( seq )
end
- # raises ArgumentError, StandardError
+ def process_hmmscan_datas( hmmscan_datas,
+ in_msa,
+ add_position,
+ add_domain_number,
+ add_species,
+ out_msa,
+ out_msa_singles,
+ out_msa_pairs,
+ out_msa_isolated,
+ min_linker )
+
+ actual_out_of = hmmscan_datas.size
+
+ hmmscan_datas.each_with_index do |hmmscan_data, index|
+ if hmmscan_data.number < ( index + 1 )
+ error_msg = "hmmscan_data.number < ( index + 1 ) (this should not have happened)"
+ raise StandardError, error_msg
+ end
+
+ extract_domain( hmmscan_data.seq_name,
+ index + 1,
+ actual_out_of,
+ hmmscan_data.env_from,
+ hmmscan_data.env_to,
+ in_msa,
+ out_msa,
+ add_position,
+ add_domain_number,
+ add_species )
+
+ if min_linker
+ if actual_out_of == 1
+ extract_domain( hmmscan_data.seq_name,
+ 1,
+ 1,
+ hmmscan_data.env_from,
+ hmmscan_data.env_to,
+ in_msa,
+ out_msa_singles,
+ add_position,
+ add_domain_number,
+ add_species )
+ else
+ first = index == 0
+ last = index == hmmscan_datas.length - 1
+
+ if ( ( first && ( ( hmmscan_datas[ index + 1 ].env_from - hmmscan_data.env_to ) > min_linker) ) ||
+ ( last && ( ( hmmscan_data.env_from - hmmscan_datas[ index - 1 ].env_to ) > min_linker ) ) ||
+ ( !first && !last && ( ( hmmscan_datas[ index + 1 ].env_from - hmmscan_data.env_to ) > min_linker ) &&
+ ( ( hmmscan_data.env_from - hmmscan_datas[ index - 1 ].env_to ) > min_linker ) ) )
+
+ extract_domain( hmmscan_data.seq_name,
+ index + 1,
+ actual_out_of,
+ hmmscan_data.env_from,
+ hmmscan_data.env_to,
+ in_msa,
+ out_msa_isolated,
+ add_position,
+ add_domain_number,
+ add_species )
+
+ elsif !first
+ extract_domain( hmmscan_data.seq_name,
+ index.to_s + "+" + ( index + 1 ).to_s,
+ actual_out_of,
+ hmmscan_datas[ index - 1 ].env_from,
+ hmmscan_data.env_to,
+ in_msa,
+ out_msa_pairs,
+ add_position,
+ add_domain_number,
+ add_species )
+ end
+ end
+ end
+ end # each
+ end # def process_hmmscan_data
+
def extract_domain( sequence,
number,
out_of,
out_msa,
add_position,
add_domain_number,
- add_domain_number_as_digit,
- add_domain_number_as_letter,
- trim_name,
add_species )
if number.is_a?( Fixnum ) && ( number < 1 || out_of < 1 || number > out_of )
error_msg = "impossible: number=" + number.to_s + ", out of=" + out_of.to_s
- raise ArgumentError, error_msg
+ raise StandardError, error_msg
end
if seq_from < 1 || seq_to < 1 || seq_from >= seq_to
error_msg = "impossible: seq-f=" + seq_from.to_s + ", seq-t=" + seq_to.to_s
- raise ArgumentError, error_msg
+ raise StandardError, error_msg
end
seqs = in_msa.find_by_name_start( sequence, true )
if seqs.length < 1
error_msg = "sequence \"" + sequence + "\" not found in sequence file"
- raise StandardError, error_msg
+ raise IOError, error_msg
end
if seqs.length > 1
error_msg = "sequence \"" + sequence + "\" not unique in sequence file"
- raise StandardError, error_msg
+ raise IOError, error_msg
end
# hmmsearch is 1 based, wheres sequences are 0 bases in this package.
seq = in_msa.get_sequence( seqs[ 0 ] ).get_subsequence( seq_from - 1, seq_to - 1 )
seq.set_name( seq.get_name + "_" + seq_from.to_s + "-" + seq_to.to_s )
end
- if trim_name
- seq.set_name( seq.get_name[ 0, seq.get_name.length - TRIM_BY ] )
+ if out_of != 1 && add_domain_number
+ seq.set_name( seq.get_name + "~" + number.to_s + "-" + out_of.to_s )
end
- if out_of != 1
- if add_domain_number_as_digit
- seq.set_name( seq.get_name + number.to_s )
- elsif add_domain_number_as_letter
- if number > 25
- error_msg = 'too many identical domains per sequence, cannot use letters to distinguish them'
- raise StandardError, error_msg
- end
- seq.set_name( seq.get_name + ( number + 96 ).chr )
- elsif add_domain_number
- seq.set_name( seq.get_name + "~" + number.to_s + "-" + out_of.to_s )
- end
- end
-
- # if ( seq.get_name.length > 10 )
- # error_msg = "sequence name [" + seq.get_name + "] is longer than 10 characters"
- # raise StandardError, error_msg
- # end
-
if add_species
a = orig_name.rindex "["
b = orig_name.rindex "]"
end # class HmmscanDomainExtractor
+ class HmmsearchData
+ def initialize( seq_name, number, out_of, env_from, env_to, i_e_value )
+ @seq_name = seq_name
+ @number = number
+ @out_of = out_of
+ @env_from = env_from
+ @env_to = env_to
+ @i_e_value = i_e_value
+ end
+ attr_reader :seq_name, :number, :out_of, :env_from, :env_to, :i_e_value
+ end
+
end # module Evoruby