File Coverage

blib/lib/TIGR/HmmTools.pm

Criterion	Covered	Total	%
statement	17	422	4.0
branch	0	190	0.0
condition	0	33	0.0
subroutine	6	14	42.8
pod	0	7	0.0
total	23	666	3.4

package TIGR::HmmTools;
require Exporter;
use 5.006;
use strict;
use warnings;

=head1 NAME

TIGR::HmmTools - Functions for processing files from HMMer

=head1 VERSION

Version 0.01

=cut

our $VERSION = '0.03';


=head1 SYNOPSIS

Ages ago developers at The Institute for Genomic Research (RIP) produced code to process
raw alignment files from the HMMer suite.  Many of these functions were embedded within a
variety of scripts which do things like format conversion.  Installing this package also
adds the utility script, 'convert_hmmscan_to_htab.pl'

=head1 EXPORT

A list of functions that can be exported.  You can delete this section
if you don't export anything, such as for a purely object-oriented module.
The following functions are exported:

    - read_hmmer3_output
    - read_hmmer2_output
    - print_htab
    - hmm_database_info
    - get_cutoffs_for_hmm_accession
    - build_alignment
    - read_hmmer3_output2

=head1 AUTHOR

Joshua Orvis, C<<  >>

The actual author of the code is currently unknown.  I just packaged it for CPAN.

=head1 BUGS

Please report any bugs or feature requests to C, or through
the web interface at L.  I will be notified, and then you'll
automatically be notified of progress on your bug as I make changes.

=head1 SUPPORT

You can find documentation for this module with the perldoc command.

    perldoc TIGR::HmmTools


You can also look for information at:

=over 4

=item * RT: CPAN's request tracker (report bugs here)

L

=item * AnnoCPAN: Annotated CPAN documentation

L

=item * CPAN Ratings

L

=item * Search CPAN

L

=back


=head1 ACKNOWLEDGEMENTS


=head1 LICENSE AND COPYRIGHT

Copyright 2017 Joshua Orvis.

This program is free software; you can redistribute it and/or modify it
under the terms of the the Artistic License (2.0). You may obtain a
copy of the full license at:

L

Any use, modification, and distribution of the Standard or Modified
Versions is governed by this Artistic License. By using, modifying or
distributing the Package, you accept this license. Do not use, modify,
or distribute the Package, if you do not accept this license.

If your Modified Version has been derived from a Modified Version made
by someone other than you, you are nevertheless required to ensure that
your Modified Version complies with the requirements of this license.

This license does not grant you the right to use any trademark, service
mark, tradename, or logo of the Copyright Holder.

This license includes the non-exclusive, worldwide, free-of-charge
patent license to make, have made, use, offer to sell, sell, import and
otherwise transfer the Package with respect to any patent claims
licensable by the Copyright Holder that are necessarily infringed by the
Package. If you institute patent litigation (including a cross-claim or
counterclaim) against any party alleging that the Package constitutes
direct or contributory patent infringement, then this Artistic License
to you shall terminate on the date that such litigation is filed.

Disclaimer of Warranty: THE PACKAGE IS PROVIDED BY THE COPYRIGHT HOLDER
AND CONTRIBUTORS "AS IS' AND WITHOUT ANY EXPRESS OR IMPLIED WARRANTIES.
THE IMPLIED WARRANTIES OF MERCHANTABILITY, FITNESS FOR A PARTICULAR
PURPOSE, OR NON-INFRINGEMENT ARE DISCLAIMED TO THE EXTENT PERMITTED BY
YOUR LOCAL LAW. UNLESS REQUIRED BY LAW, NO COPYRIGHT HOLDER OR
CONTRIBUTOR WILL BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, OR
CONSEQUENTIAL DAMAGES ARISING IN ANY WAY OUT OF THE USE OF THE PACKAGE,
EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.


=cut

use Carp;
use Data::Dumper;
our @ISA    = qw (Exporter);
our @EXPORT =
  qw(read_hmmer3_output read_hmmer2_output print_htab hmm_database_info get_cutoffs_for_hmm_accession build_alignment read_hmmer3_output2);
our @EXPORT_OK = qw ();

sub read_hmmer3_output {
    my ($path) = @_;
    my $retval = {};

    my $in_result = 0;
    my $in_hit_scores = 0;
    my $in_domain_scores = 0;

    if ( $path ne '' ) {
	chomp $path;
	my @statd = stat $path;
	$retval->{'info'}->{ 'search_date' } = ( ( localtime( $statd[ 9 ] ) )[ 3 ] ) . "-"
	    . ( ( localtime( $statd[ 9 ] ) )[ 4 ] + 1 ) . "-"
	    . ( ( localtime( $statd[ 9 ] ) )[ 5 ] + 1900 );
    } else {
	$retval->{'info'}->{ 'search_date' } = ( ( localtime )[ 3 ] ) . "-"
	    . ( ( localtime )[ 4 ] + 1 ) . "-"
	    . ( ( localtime )[ 5 ] + 1900 );
    }

    open(my $fh, "< $path") or die("Unable to open $path: $!");
    while( my $line = <$fh> ) {
	next if( $line =~ /^\s*$/ || $line =~ /inclusion threshold/ );

	# The program line
	if ( $line =~ /^\#\s*((hmmscan)\.*)/ ) {
	    $retval->{'info'}->{'program'} = $1;

	 # The version
	 } elsif( $line =~ /^\#\s*HMMER ([\d\.]+)\s+\(([^\)]+)\)/ ) {
	     ( $retval->{'info'}->{'version'}, $retval->{'info'}->{'release'} ) = ( $1, $2 );

	 # The hmm database searched
	 } elsif( $line =~ /^\#\s+target HMM database:\s+(\S+)/ ) {
	     $retval->{'info'}->{'hmm_file'} = $1;

	 # The query file
	 } elsif( $line =~ /^\#\squery sequence (file|database):\s+(.+)/ ) {
	     $retval->{'info'}->{'sequence_file'} = $2;

	 # This indicates we're parsing a hit.
	 } elsif( $line =~ /^Query\:\s+(\S+)/ ) {
	     my $data = &_parse_hmmpfam3_hit( $fh );
	     $retval->{'queries'}->{$1} = $data;

	 }
    }

    close($fh);
    return $retval;
}

sub _parse_hmmpfam3_hit {
    my ($fh) = @_;
    my $data = {};

    my $in_hit_scores = 0;
    my $in_domain_scores = 0;
    my $hit_acc;

    while( my $line = <$fh> ) {
	chomp($line);
	last if( $line =~ m|//| );
	next if( $line =~ /^\s*$/ );

	if( $line =~ /^Domain annotation/ ) {
	    $in_hit_scores = 0;

	} elsif( $in_hit_scores && $line !~ /^\s+--/ && $line !~ /inclusion_threshold/ ) {
	    if( $line =~ /No hits detected/ ) {
		$data->{'hits'} = {};
		last;
	    }

	    my @c = split(/\s+/, $line);
	    my $t_hit_acc = $c[9];
	    $data->{'hits'}->{$t_hit_acc}->{total_evalue} = $c[1];
	    $data->{'hits'}->{$t_hit_acc}->{total_score} = $c[2];
	    $data->{'hits'}->{$t_hit_acc}->{accession} = $c[9];
	    $data->{'hits'}->{$t_hit_acc}->{hit_description} = join(" ",@c[10..(@c-1)]);
	    $data->{'hits'}->{$t_hit_acc}->{domain_count} = $c[8];
	    $data->{'hits'}->{$t_hit_acc}->{frame} = "";
	} elsif( $line =~ /^>>\s*(\S+)/ ) {
	    $in_domain_scores = 1;
	    $hit_acc = $1;

	} elsif( $in_domain_scores && $line =~ /Alignments for each/ ) {
	    undef $hit_acc;
	    $in_domain_scores = 0;

	} elsif( $in_domain_scores && $line !~ /^\s*[>\#-]/ ) {
	    die("Didn't parse hit accession from header line before getting to domain table")
		unless( $hit_acc );
	    die("The hit accession [$hit_acc] didn't exist in lookup") unless( exists( $data->{'hits'}->{$hit_acc} ) );

	    if( $line =~ /\[No individual domains that/ ) {
		$data->{'hits'}->{$hit_acc}->{'domains'} = {};
		$in_domain_scores = 0;
		next;
	    }

	    my @c = split(/\s+/, $line);

	    if( $c[1] eq 'targets' ||  $c[1] eq 'reported' || $c[1] eq 'Fwd') {
		print Dumper( $data->{'hits'}->{$hit_acc} );
		print "LINE: $line\n";
		print Dumper( @c );
		die("Issue parsing");
	    }

	    my $a = $c[1];
	    my $b = $a + 0;
	    if( $b ne $a ) {
		print Dumper( $data->{'hits'}->{$hit_acc} );
		print "LINE: $line\n";
		print Dumper( \@c );
		die("c[1] not numeric");
	    }


	    $data->{'hits'}->{$hit_acc}->{'domains'}->{$c[1]}->{'seq_f'} = $c[10];
	    $data->{'hits'}->{$hit_acc}->{'domains'}->{$c[1]}->{'seq_t'} = $c[11];
	    $data->{'hits'}->{$hit_acc}->{'domains'}->{$c[1]}->{'hmm_f'} = $c[7];
	    $data->{'hits'}->{$hit_acc}->{'domains'}->{$c[1]}->{'hmm_t'} = $c[8];
	    $data->{'hits'}->{$hit_acc}->{'domains'}->{$c[1]}->{'domain_score'} = $c[3];
	    $data->{'hits'}->{$hit_acc}->{'domains'}->{$c[1]}->{'domain_evalue'} = $c[6];
	} elsif( $line =~ /^\s+E-value\s+score/ ) {
	    $in_hit_scores = 1;
	}
    }

    return $data;
}

## Subroutine to parse hmmscan (Hmmer3.0) output
sub read_hmmer3_output_old {
    my $path = shift;
    my $data = {};
    my @lines;
    if ( $path ne '' ) {
	chomp $path;
	my @statd = stat $path;
	$data->{ 'search_date' } = ( ( localtime( $statd[ 9 ] ) )[ 3 ] ) . "-"
	    . ( ( localtime( $statd[ 9 ] ) )[ 4 ] + 1 ) . "-"
	    . ( ( localtime( $statd[ 9 ] ) )[ 5 ] + 1900 );
	open( FH, "$path" ) || die "Can't open $path for reading: $!\n";
	chomp( @lines =  );
	close FH;
    } else {
	chomp( @lines =  );
	$data->{ 'search_date' } = ( ( localtime )[ 3 ] ) . "-"
	    . ( ( localtime )[ 4 ] + 1 ) . "-"
	    . ( ( localtime )[ 5 ] + 1900 );
    }
    if ( !@lines ) {
	carp "No data read from input $path";
	return undef;
    }
    my $i = 0;
    while ($i < @lines) {
	if ( $lines[ $i ] =~ /^#\s*((hmmscan)\.*)/ ) {
	     $data->{ 'program' } = $1;
	     my $version_line = $lines[ ++$i ];
	     $version_line =~ /^#\s*HMMER (\d+)\.(\S+)/;
	     ( $data->{ 'version' }, $data->{ 'release' } ) = ( $1, $2 );
	     $i +=2;
	     last;
	 }
	++$i;
    }
    until ( $lines[ $i ] =~ /^\s*$/ ) {
	if ( $lines[ $i ] =~ /^#\s*target HMM database:\s+(\S+)/ ) {
	     $data->{ 'hmm_file' } = $1;
	 } elsif ( $lines[ $i ] =~ /^#\s*query sequence (file|database):\s+(.+)/ ) {
		   $data->{ 'sequence_file' } = $2;
	       }
	$i++;
	die "Failure to parse" if $i > @lines;
    }
    $i++;
    if ( $lines[ $i ] =~ /^Query:\s+(\S+)/ ) {
	$data->{ 'query' } = $1;
	print "Found query: $1\n";
	$i++;
    }
    until ( $lines[$i] =~ /^\s*$/ ) {
	if ( $lines[ $i ] =~ /^Scores for/ ) {
	    $i++; # this skips the separator row
	    my $headers = $lines[ ++$i ];
	    $i++;        # this skips the separator row
	} elsif ( $lines[ $i ] !~ /No hits detected that satisfy reporting thresholds/i ) {
	    $i++ if ($lines[$i] =~ /inclusion threshold/g);
	    my @c = split /\s+/, $lines[ $i ],11;
	    my $hit_index = $c[9];
	    $data->{hit}{$hit_index}{total_evalue} = $c[1];
	    $data->{hit}{$hit_index}{total_score} = $c[2];
	    $data->{hit}{$hit_index}{accession} = $c[9];
	    $data->{hit}{$hit_index}{hit_description} = $c[10];
	    $data->{hit}{$hit_index}{domain_count} = $c[8];
	    $data->{hit}{$hit_index}{frame} = "";
	} else {
	    return $data;
	}
	$i++;
	die "Failure to parse" if $i > @lines;
    }
    $i++;
    if ( $lines[ $i ] =~ /^Domain annotation for each model/ ) {
	$i++;
    }
    until ($lines[$i] =~ /^\/\/$/) {
	if ( $lines[ $i ] !~ /No targets detected that satisfy reporting thresholds/ ) {
	    if($lines[$i] =~ />>/) {
		my @c = split /\s+/, $lines[ $i ];
		my $hit_index = $c[ 1 ];
		$i += 3;

		if ( !defined $data->{ 'hit' }->{ $hit_index } ) {
		    warn "Why doesn't '$hit_index' match an existing identifier?";
		} else {
		    until ($lines[$i] =~ /^\s*$/) {
			my @res = split /\s+/, $lines[ $i ];
			$data->{hit}{$hit_index}{domain}{$res[1]}{seq_f} = $res[10];
			$data->{hit}{$hit_index}{domain}{$res[1]}{seq_t} = $res[11];
			$data->{hit}{$hit_index}{domain}{$res[1]}{hmm_f} = $res[7];
			$data->{hit}{$hit_index}{domain}{$res[1]}{hmm_t} = $res[8];
			$data->{hit}{$hit_index}{domain}{$res[1]}{domain_score} = $res[3];
			$data->{hit}{$hit_index}{domain}{$res[1]}{domain_evalue} = $res[6];
			$i++;
		    }
		}
	    }
	} else {
	    return $data;
	}
	$i++;
    }
    return $data;
}
## Use with caution. Or just don't use this method.
sub read_hmmer2_output {
	my $path = shift;
	my $data = {};
# die "Bad Parser under reconstruction.\n\n";
# drink in data
	my @lines;

# drink in the output from file or stdin
	if ( $path ne '' ) {
		chomp $path;
		my @statd = stat $path;
		$data->{ 'search_date' } =
			( ( localtime( $statd[ 9 ] ) )[ 3 ] ) . "-"
			. ( ( localtime( $statd[ 9 ] ) )[ 4 ] + 1 ) . "-"
			. ( ( localtime( $statd[ 9 ] ) )[ 5 ] + 1900 );
		open( FH, "$path" )
			|| die "Can't open $path for reading: $!\n";
		chomp( @lines =  );
		close FH;
	}
	else {
		chomp( @lines =  );
		$data->{ 'search_date' } =
			( ( localtime )[ 3 ] ) . "-"
			. ( ( localtime )[ 4 ] + 1 ) . "-"
			. ( ( localtime )[ 5 ] + 1900 );
	}
	if ( !@lines ) {
		carp "No data read from input $path";
		return undef;
	}
	my $i = 0;

# first line grouping is company, package and license info
# warn "Parsing License. Line $i\n";
# amahurkar:1/15/08 Seems like the current output does not have licesning info, so commenting this
# so commenting out these lines
#until ( $lines[ $i ] eq "" )
#    $data->{ 'header' } .= $lines[ $i ] . "\n";
#    $i++;

#$i++;

# next group is program and version
# warn "Parsing Program and version. Line $i\n";
# amahurkar:1/15/08 the format has changed and now there is no blank space
# after program name, so we are using '- -' as tha match param
# to stop parsing for program name
#until ( $lines[ $i ] eq "" )
#until ( $lines[ $i ] =~ m/^-\s-/)  # It doesn't work with LDhmmpfam v1.5.4
#until ( $lines[ $i ] =~ m/^-{8}\s+-/)
	while ($i < @lines) {
		if ( $lines[ $i ] =~ /^((hmmpfam|hmmsearch)\.*)/ ) {
			$data->{ 'program' } = $1;
			my $version_line = $lines[ ++$i ];
			$version_line =~ /^HMMER (\d+)\.(\S+)/;
			( $data->{ 'version' }, $data->{ 'release' } ) = ( $1, $2 );
			$i +=2;
			last;
		}
		++$i;
	}

# next group is program parameters
# warn "Parsing Parameters. Line $i\n";
# amahurkar:1/15/08 the format has changed and now there is no blank space
# after program name, so we are using '- -' as tha match param
# to stop parsing for program parameters
	until ( $lines[ $i ] =~ /^\s*$/ ) {
		if ( $lines[ $i ] =~ /^HMM file:\s+(\S+)/ ) {
			$data->{ 'hmm_file' } = $1;
		} elsif ( $lines[ $i ] =~ /^Sequence (file|database):\s+(.+)/ ) {
			$data->{ 'sequence_file' } = $2;
		} elsif ( $lines[ $i ] =~ /^per-sequence score cutoff:\s+(.+)/ ) {
			$data->{ 'total_score_cutoff' } = $1;
		} elsif ( $lines[ $i ] =~ /^per-domain score cutoff:\s+(.+)/ ) {
			$data->{ 'domain_score_cutoff' } = $1;
		} elsif ( $lines[ $i ] =~ /^per-sequence E-value cutoff:\s+(.+)/ ) {
			$data->{ 'total_evalue_cutoff' } = $1;
		} elsif ( $lines[ $i ] =~ /^per-domain E-value cutoff:\s+(.+)/ ) {
			$data->{ 'domain_evalue_cutoff' } = $1;
		}
		$i++;
		die "Failure to parse" if $i > @lines;
	}
	$i++;

# get query info
# warn "Parsing Query Info. Line $i\n";
#until ( $lines[ $i ] eq "" )
	until ( $lines[$i] =~ /^\s*$/ ) {
		if ( $lines[ $i ] =~ /^Query (?:HMM|sequence):\s+(.+)/ ) {
			$data->{ 'query' } = $1;
		}
		elsif ( $lines[ $i ] =~ /^Accession:\s+(.+)/ ) {
			$data->{ 'query_accession' } = $1;
		}
		elsif ( $lines[ $i ] =~ /^Description:\s+(.+)/ ) {
			$data->{ 'query_description' } = $1;
		}
		$i++;
		die "Failure to parse" if $i > @lines;
	}
	$i++;

# next section is global search results
# warn "Parsing Global Search Results. Line $i\n";
	my $find_frame = 0;  # is datbase nucleotide sequence?
	my $hit_index;
	until ( $lines[$i] =~ /^\s*$/ ) {
		if ( $lines[ $i ] =~ /^Scores for/ ) {
			my $headers = $lines[ ++$i ];
			if ( $headers =~ /\bFr\b/ ) {
				$data->{ 'program' } .= "-frames";
				$find_frame = 1;
			}
			$i++;        # this skips the separator row
		} elsif ( $lines[ $i ] !~ /no hits above thresholds/i ) {
			my @c = split /\s+/, $lines[ $i ];
			if ( $find_frame ) {
				$hit_index = $c[ 0 ] . $c[ $#c ];
				$data->{ 'hit' }->{ $hit_index }->{ 'frame' } = pop @c;
			}
			else {
# $hit_index = $c[ 0 ]; # AP 20090807
				($hit_index = $c[0]) =~ s/\.\d+$//;
			}
			$data->{hit}{$hit_index}{accession}       = shift @c;
			$data->{hit}{$hit_index}{domain_count}    = pop @c;
			$data->{hit}{$hit_index}{total_evalue}    = pop @c;
			$data->{hit}{$hit_index}{total_score}     = pop @c;
			$data->{hit}{$hit_index}{hit_description} = join " ", @c;
		}
		else {
			return $data;
		}
		$i++;
		die "Failure to parse" if $i > @lines;
	}
	$i++;

# next section is domain breakdown
# warn "Parsing Domain Breakdown. Line $i\n";
#until ( $lines[ $i ] eq "" )
	until ($lines[$i] =~ /^\s*$/) {
		if ( $lines[ $i ] =~ /^Parsed for domains/ ) {
			$i += 2;  # to skip header and separator
		} elsif ( $lines[ $i ] !~ /no hits above thresholds/ ) {

# $lines[$i] =~ /^(\w+)\s+(\d+)\/\d+\s+(\d+)\s+(\d+)\s[\[\.\]]{2}\s+(\d+)\s+(\d+)\s[\[\.\]]{2}\s+(-?[\.\d]+)\s+([\.\-e\d]+)\s*(\-?\d)?/;
			my @c = split /\s+/, $lines[ $i ];
			if ( $find_frame ) {
				$hit_index = $c[ 0 ] . $c[ $#c ];
			}
			else {
#$hit_index = $c[ 0 ]; # AP 20090807
				($hit_index = $c[0]) =~ s/\.\d+$//;
			}
			if ( !defined $data->{ 'hit' }->{ $hit_index } ) {
				warn
					"Why doesn't '$hit_index' match an existing identifier?";
			}

# $data->{'hit'}->{$hit_index}->{'domain'}->{$2}->{'seq_f'} = $3;
			my ( $d, $t ) = split /\//, $c[ 1 ];
			$data->{hit}{$hit_index}{domain}{$d}{seq_f} = $c[2];
			$data->{hit}{$hit_index}{domain}{$d}{seq_t} = $c[3];
			$data->{hit}{$hit_index}{domain}{$d}{hmm_f} = $c[5];
			$data->{hit}{$hit_index}{domain}{$d}{hmm_t} = $c[6];
			$data->{hit}{$hit_index}{domain}{$d}{domain_score} = $c[8];
			$data->{hit}{$hit_index}{domain}{$d}{domain_evalue} = $c[9];
		}
		$i++;
		die "Failure to parse" if $i > @lines;
	}
	$i++;
	if ( $data->{ 'program' } =~ /hmmsearch/ ) {

# next section is alignments
# warn "Parsing Alignments. Line $i\n";
	    if ( $lines[ $i ] =~ /^Alignments of top-scoring domains/ ) {
		$i++;
		## kgalens [11/13/2012]
		## Not sure how this ever worked? $domain isn't in scope here.
		## I've added use strict; and this module wouldn't. So adding the next line
		## just so it will run.
		my $domain;
		my $hit;
		my $hit_index;
		## END kgalens [11/13/2012]

		until ( $lines[$i] =~ /^\s*$/ ) {
		    if ( $lines[ $i ] =~ /^(\S+): domain (\d+)/ ) {
			($hit_index, $hit, $domain ) = ( $1, $1, $2 );
			$hit =~ s/(.{10}).*/$1/;
# warn "  Parsing hit $hit_index. Line $i\n";
			if ( $find_frame ) {
			    if (   $lines[ $i ] =~ /Fr = ([\-\d]+)/ || $lines[ $i ] =~ /\. frame ([\-\d]+)/ ) {
				$hit_index .= $1;
			    } else {
				warn "ERROR: Couldn't find frame from:\n '$lines[$i]'";
			    }
			}
			if ( !defined $data->{ 'hit' }->{ $hit_index } ) {
			    warn "Why doesn't '$hit_index' match an existing identifier?";
			}
			$i++;
		    }
		    if ( $lines[ $i ] =~ /\bRF\b/ ) { ##  WHY??
			$i++;
		    }

# capture aligned hmm consensus
		    my $hmm_seq = $lines[ $i ];
		    $hmm_seq =~ s/\s+//g;
		    $hmm_seq =~ s/[\*\-\>\<]//g;

		    $data->{ 'hit' }->{ $hit_index }->{ 'domain' }->{ $domain }->{hmm_seq} .= $hmm_seq;
		    until ( $lines[ $i ] =~ /^\s+\Q$hit\E/) { ## changed from /\b\Q$hit\E\b/ )
			$i++;
		    }
		    my $prot_seq = $lines[ $i ];
		    if ( $prot_seq =~ /\w+\s+(\d+|\-)\s+(\S+)\s+(\d+|\-)/ ) {
			$data->{ 'hit' }->{ $hit_index }->{ 'domain' }->{ $domain }->{prot_seq} .= $2;
		    }
		    $i += 2;  # skip the blank line and move on to the next
		    die "Failure to parse" if $i > @lines;
		}
	    }
	    $i++;

# next (last) section is statistics
# warn "Parsing Statistics. Line $i\n";
	    my @data;
	    while ( $i < @data ) {
		if ( $lines[ $i ] =~ /^\s+mu =\s+(-?\d+)/ ) {
		    $data->{ 'mu' } = $1;
		} elsif ( $lines[ $i ] =~ /^\s+lambda =\s(-?\d+)/ ) {
		    $data->{ 'lambda' } = $1;
		} elsif ( $lines[ $i ] =~ /chi-sq statistic =\s(\d+)/ ) {
		    $data->{ 'chisq' } = $1;
		} elsif ( $lines[ $i ] =~ /Total sequences searched:\s*(\d+)/ ) {
		    $data->{ 'tot_seq_searched' } = $1;
		} elsif ( $lines[ $i ] =~ /Whole sequence top hits/ ) {
		    $lines[ ++$i ] =~ /(\d+)/;
		    $data->{ 'total_hits' } = $1;
		    $lines[ ++$i ] =~ /(\d+)/;
		    $data->{ 'total_hits_above_evalue_cutoff' } = $1;
		} elsif ( $lines[ $i ] =~ /Domain top hits/ ) {
		    $lines[ ++$i ] =~ /(\d+)/;
		    $data->{ 'domain_hits' } = $1;
		    $lines[ ++$i ] =~ /(\d+)/;
		    $data->{ 'domain_hits_above_evalue_cutoff' } = $1;
		}
		$i++;
		die "Failure to parse" if $i > @lines;
	    }
	}
	return $data;
}

sub hmm_database_info {
    my $dbh   = shift;
    my $hmm_q =
      "SELECT hmm_acc, hmm_len, trusted_cutoff, noise_cutoff, hmm_com_name,"
      . " trusted_cutoff2, noise_cutoff2, gathering_cutoff, gathering_cutoff2"
      . " FROM hmm2"
      . " WHERE is_current = 1";
    my $HMM = $dbh->selectall_hashref( $hmm_q, 'hmm_acc' );
    return $HMM;
}

sub print_htab {
#NOTE:  This will produce results if the data hash was created through the 'read_hmmer3_output' subroutine
#		This is because of a change in the naming of some of the property keys (ex 'hit' to 'hits')
    my $data   = shift;
    my $HMM    = shift;
    my $output = shift;
    foreach my $qry_id ( keys %{$data->{'queries'}} ) {
    	foreach my $hit (
        	sort {
                    $data->{'queries'}->{$qry_id}->{'hits'}->{ $b }->{ 'total_score' } <=> $data->{'queries'}->{$qry_id}->
                    	{'hits'}->{ $a }->{ 'total_score' }
        	} keys %{ $data->{'queries'}->{$qry_id}->{'hits'} } )
	{
            my $h = $data->{'queries'}->{$qry_id}->{'hits'}->{ $hit };
            next if (scalar keys %{$h->{'domains'}} == 0);	#skip model hits that have no domain hits
            foreach my $domain ( sort { $a <=> $b } keys %{ $h->{ 'domains' } } )
            {
            	# for convenience
            	my $dh = $h->{ 'domains' }->{ $domain };
            	if ( $data->{'info'}->{ 'program' } =~ /hmmsearch/ ) {	#hmmer2 is currently deprecated so this will probably error
                	my $hmm_com_name =
                	    $HMM->{ $data->{ 'query' } }->{ 'hmm_com_name' }
               	   		? $HMM->{ $data->{ 'query' } }->{ 'hmm_com_name' }
                  		: $data->{ 'query_description' };
                	print $output "$data->{query}"
                  	    . "\t$data->{search_date}"
              		    . "\t$HMM->{$data->{query}}->{hmm_len}"
                	    . "\t$data->{program}"
                 	    . "\t$data->{sequence_file}"
             		    . "\t$h->{accession}"
             		    . "\t$dh->{hmm_f}"
             		    . "\t$dh->{hmm_t}"
              		    . "\t$dh->{seq_f}"
             		    . "\t$dh->{seq_t}"
                	    . "\t$h->{frame}"
               		    . "\t$dh->{domain_score}"
               		    . "\t$h->{total_score}"
                	    . "\t$domain"
                	    . "\t$h->{domain_count}"
               		    . "\t$hmm_com_name"
               		    . "\t$h->{hit_description}"
                	    . "\t$HMM->{$data->{query}}->{trusted_cutoff}"
                	    . "\t$HMM->{$data->{query}}->{noise_cutoff}"
                	    . "\t$h->{total_evalue}"
                	    . "\t$dh->{domain_evalue}"
                	    . "\t$HMM->{$data->{query}}->{trusted_cutoff2}"
                	    . "\t$HMM->{$data->{query}}->{noise_cutoff2}"
                	    . "\t$HMM->{$data->{query}}->{gathering_cutoff}"
                	    . "\t$HMM->{$data->{query}}->{gathering_cutoff2}" . "\n";
            	}
            	elsif ( $data->{'info'}->{ 'program' } =~ /hmmscan|hmmpfam/ ) {
                    my $hmm_com_name =
                	    $HMM->{ $hit }->{ 'hmm_com_name' }
                	  ? $HMM->{ $hit }->{ 'hmm_com_name' }
                	  : $h->{ 'hit_description' };
                    print $output "$h->{accession}"
                	  . "\t$data->{'info'}->{search_date}"
                	  . "\t$HMM->{$hit}->{hmm_len}"
                	  . "\t$data->{'info'}->{program}"
                	  . "\t$data->{'info'}->{hmm_file}"
                 	  . "\t$qry_id"
                	  . "\t$dh->{hmm_f}"
                 	  . "\t$dh->{hmm_t}"
                  	  . "\t$dh->{seq_f}"
                  	  . "\t$dh->{seq_t}"
               		  . "\t$h->{frame}"
                	  . "\t$dh->{domain_score}"
                	  . "\t$h->{total_score}"
                	  . "\t$domain"
                	  . "\t$h->{domain_count}"
                	  . "\t$hmm_com_name"
                	  . "\t$h->{hit_description}"
                	  . "\t$HMM->{$h->{accession}}->{trusted_cutoff}"
                	  . "\t$HMM->{$h->{accession}}->{noise_cutoff}"
                	  . "\t$h->{total_evalue}"
                	  . "\t$dh->{domain_evalue}"
                	  . "\t$HMM->{$h->{accession}}->{trusted_cutoff2}"
                	  . "\t$HMM->{$h->{accession}}->{noise_cutoff2}"
                	  . "\t$HMM->{$h->{accession}}->{gathering_cutoff}"
                	  . "\t$HMM->{$h->{accession}}->{gathering_cutoff2}" . "\n";
            	}
            }
        }
    }
}

sub build_alignment {
    my $data         = shift;
    my $instructions = shift;

    # build output file name
    my $output_file;
    $output_file =
      $instructions->{file_prefix} . "." . $instructions->{file_format};
    open my $OUT, ">$output_file"
      or croak "Can't open '$output_file' as output file: $!\n";
    select $OUT;

    # retrieve aligned sequences
    my %screened;
    foreach my $hit ( keys %{ $data->{ 'hit' } } ) {

        # screen for total score cutoffs
        if ( $data->{hit}->{ $hit }->{total_score} >=
               $instructions->{total_bit_cutoff}
            && $data->{hit}->{ $hit }->{total_evalue} <=
            $instructions->{total_evalue_cutoff} )
        {
            foreach my $domain ( keys %{ $data->{hit}->{ $hit }->{domain} } )
            {
                if ( $data->{hit}->{ $hit }->{domain}->{ $domain }
                    ->{domain_score} >= $instructions->{domain_bit_cutoff}
                    && $data->{hit}->{ $hit }->{domain}->{ $domain }
                    ->{domain_evalue} <=
                    $instructions->{domain_evalue_cutoff} )
                {
                    $screened{ $hit } = $domain;
                }
            }
        }
    }

    # Now that we have sequences aligned to hmm sequence, we have to translate
    # this into a multiple alignment. Assign each position in each alignment to
    # a position on the hmm 'sequence', and keep track of gaps in the hmm alignment
    my %DIST;
    my $ref;
    foreach my $hit ( keys %screened ) {
        $ref =
          $data->{ 'hit' }->{ $hit }->{ 'domain' }->{ $screened{ $hit } };

        # split aligned hmm seq and aligned protein seq into arrays
        my @hmma  = split / */, $ref->{hmm_seq};
        my @prota = split / */, $ref->{prot_seq};

        # these should be the same length. If not, there's an error.
        if ( @hmma != @prota ) {
            croak "Length of hmm alignment (" . @hmma . ")"
              . " is not equal to protein alignment ("
              . @prota . ")"
              . ": $data->{query}/$data->{hit}->{$hit}->{accession}\n"
              . "$ref->{hmm_seq}\n@hmma\n$ref->{prot_seq}\n@prota\n";
        }

       # assign each position in the protein alignment to its hmm alignment position,
       # if one exists
        my $hmm_pos = $ref->{hmm_f};
        my $gap     = 0;
        for ( my $i = 0 ; $i < @hmma ; $i++ ) {
            if ( $hmma[ $i ] ne "." ) {

                # assign position in the protein alignment to its hmm alignment position,
                $prota[ $i ] = $hmm_pos;

                # record max gap distance between hmm alignment positions.
                $DIST{ $hmm_pos } = $gap
                  if ( $gap >= $DIST{ $hmm_pos } );
                $gap = 0;
                $hmm_pos++;
            }
            else {
                $gap++;
            }
        }
        $ref->{aln_map} = \@prota;
    }

    # Now go back through (now that we've fully expanded the hmm alignment
    # to include any and all gaps) and make aligned protein sequence
    foreach my $hit ( keys %screened ) {
        $ref =
          $data->{ 'hit' }->{ $hit }->{ 'domain' }->{ $screened{ $hit } };
        my @prot_seq = split / */, $ref->{prot_seq};

        # start our aligned protein with any gap resulting from a partial HMM hit.
        my $aln_prot = "." x ( $ref->{hmm_f} - 1 );
        my $insert = 0;
        for ( my $i = 0 ; $i < @prot_seq ; $i++ ) {

            # grab the hmm alignment position for each protein alignment position
            my $pos = $ref->{aln_map}->[ $i ];
            if ( $pos =~ /\d+/ ) {

                # if it maps to a position, first insert any gap from the hmm alignment
                $aln_prot .= "." x ( $DIST{ $pos } - $insert );

                # then add the aa.
                $aln_prot .= $prot_seq[ $i ];
                $insert = 0;
            }

            # if it is an insertion (ie the hmm alignment shows gap), insert a gap
            else {
                $aln_prot .= $prot_seq[ $i ];
                $insert++;
            }
        }
        $aln_prot =~ s/\-/\./g;
        $ref->{aln_prot} = $aln_prot;
    }

    # Now print out in selected format
    # Stockholm format
    if ( $instructions->{file_format} eq "mul" ) {
        print "# STOCKHOLM 1.0\n";
        foreach my $hit (
            sort {
                $data->{ 'hit' }->{ $b }
                  ->{ 'total_score' } <=> $data->{ 'hit' }->{ $a }
                  ->{ 'total_score' }
            } keys %screened
          )
        {
            my $domain     = $screened{ $hit };
            my $hit_ref    = $data->{hit}->{ $hit };
            my $domain_ref = $hit_ref->{domain}->{ $domain };

            # each line should look like:
            # prot_acc/coord-coord sequence
            printf "%-40s%s\n",
              (
                "$hit_ref->{accession}/$domain_ref->{seq_f}-$domain_ref->{seq_t}",
                $domain_ref->{aln_prot}
              );
        }
    }

    # FASTA format
    elsif ($instructions->{file_format} eq "fasta"
        || $instructions->{file_format} eq "fa" )
    {
        use TIGR::FASTA::Record;
        foreach my $hit (
            sort {
                $data->{ 'hit' }->{ $b }
                  ->{ 'total_score' } <=> $data->{ 'hit' }->{ $a }
                  ->{ 'total_score' }
            } keys %screened
          )
        {
            my $domain     = $screened{ $hit };
            my $hit_ref    = $data->{hit}->{ $hit };
            my $domain_ref = $hit_ref->{domain}->{ $domain };
            my $aln_prot = $domain_ref->{aln_prot};
            $aln_prot =~ s/(.{60})/$1\n/g;
            $aln_prot =~ s/\n+/\n/g;
            chomp $aln_prot;
            my $header =
              ">$hit_ref->{accession}/$domain_ref->{seq_f}-$domain_ref->{seq_t}\n";
            print $header . $aln_prot . "\n";
        }
    }

    # MSF format
    elsif ( $instructions->{file_format} eq "msf" ) {
        my $head_len = 40;
        my ( %new_acc, %tmp_seq );
        my $header_line = 0;
        my @alignment;
        foreach my $hit (
            sort {
                $data->{ 'hit' }->{ $b }
                  ->{ 'total_score' } <=> $data->{ 'hit' }->{ $a }
                  ->{ 'total_score' }
            } keys %screened
          )
        {
            my $domain     = $screened{ $hit };
            my $hit_ref    = $data->{hit}->{ $hit };
            my $domain_ref = $hit_ref->{domain}->{ $domain };
            my $len        = length( $domain_ref->{aln_prot} );
            if ( $header_line == 0 ) {

                #added DUMMY checksum for HMMER 2.2g hmmbuild. DHH.
                print
                  "PileUp\n\n   MSF:  $len  Type: P  Check: 1111  ..\n\n";
                $header_line = 1;
            }

            # print accession list at top
            printf " Name: %-40s Len:  $len  Check:   0  Weight:  1.0\n",
              "$hit_ref->{accession}/$domain_ref->{seq_f}-$domain_ref->{seq_t}";

            # prepare alignment for bottom
            my @tmp_pep = split //, $domain_ref->{aln_prot};
            for ( my $i = 0 ; $i < ( $len / 50 ) ; $i++ ) {
                $alignment[ $i ] .= sprintf "%-40s",
                  "$hit_ref->{accession}/$domain_ref->{seq_f}-$domain_ref->{seq_t}";
                for ( my $b = 0 ; $b < 5 ; $b++ ) {
                    for ( my $a = 0 ; $a < 10 ; $a++ ) {
                        $alignment[ $i ] .=
                          $tmp_pep[ $a + ( $b * 10 ) + ( 50 * $i ) ];
                    }
                    $alignment[ $i ] .= " ";
                }
                $alignment[ $i ] .= "\n";
            }
        }
        print "\n//\n\n\n";
        foreach my $block ( @alignment ) {
            print "$block\n\n";
        }
    }
    else {
        croak
          "Don't recognize alignment file format $instructions->{file_format}\n";
    }
    select STDOUT;
}

sub get_cutoffs_for_hmm_accession {
    my $dbh       = shift;
    my $accession = shift;
    my $hmm_q     =
      "select trusted_cutoff, trusted_cutoff2, noise_cutoff from egad..hmm2 where hmm_acc = '$accession'";
    return $dbh->selectrow_hashref( $hmm_q );
}

1; # End of TIGR::HmmTools