« CLIPSeqTools/App/count_reads_on_genic_elements.pm
↑ Summary
CLIPSeqTools/App/distribution_on_introns_exons.pm »
| 1 | =head1 NAME | ||
| 2 | |||
| 3 | CLIPSeqTools::App::nmer_enrichment_over_shuffled - Measure Nmer enrichment over shuffled reads. | ||
| 4 | |||
| 5 | =head1 SYNOPSIS | ||
| 6 | |||
| 7 | clipseqtools nmer_enrichment_over_shuffled [options/parameters] | ||
| 8 | |||
| 9 | =head1 DESCRIPTION | ||
| 10 | |||
| 11 | Measure words of length N (Nmer) enrichment and compare to shuffled reads. Shuffling is done at the nucleotide level and p-values are calculated using permutations. | ||
| 12 | |||
| 13 | =head1 OPTIONS | ||
| 14 | |||
| 15 | Input options for library. | ||
| 16 | --driver <Str> driver for database connection (eg. mysql, | ||
| 17 | SQLite). | ||
| 18 | --database <Str> database name or path to database file for file | ||
| 19 | based databases (eg. SQLite). | ||
| 20 | --table <Str> database table. | ||
| 21 | --host <Str> hostname for database connection. | ||
| 22 | --user <Str> username for database connection. | ||
| 23 | --password <Str> password for database connection. | ||
| 24 | --records_class <Str> type of records stored in database. | ||
| 25 | --filter <Filter> filter library. May be used multiple times. | ||
| 26 | Syntax: column_name="pattern" | ||
| 27 | e.g. keep reads with deletions AND not repeat | ||
| 28 | masked AND longer than 31 | ||
| 29 | --filter deletion="def" | ||
| 30 | --filter rmsk="undef" . | ||
| 31 | --filter query_length=">31". | ||
| 32 | Operators: >, >=, <, <=, =, !=, def, undef | ||
| 33 | |||
| 34 | Output | ||
| 35 | --o_prefix <Str> output path prefix. Script will create and add | ||
| 36 | extension to path. [Default: ./] | ||
| 37 | |||
| 38 | Other options. | ||
| 39 | --nmer_length <Int> length N of the Nmer. Default: 6 | ||
| 40 | --permutations <Int> number of permutation to be performed. | ||
| 41 | Use more than 100 to get p-values < 0.01. | ||
| 42 | --subset <Num> run analysis on random subset. Option specifies | ||
| 43 | number (if integer) or percent (if % is used) | ||
| 44 | of data to be used. | ||
| 45 | -v --verbose print progress lines and extra information. | ||
| 46 | -h -? --usage --help print help message | ||
| 47 | |||
| 48 | =cut | ||
| 49 | |||
| 50 | package CLIPSeqTools::App::nmer_enrichment_over_shuffled; | ||
| 51 | $CLIPSeqTools::App::nmer_enrichment_over_shuffled::VERSION = '1.0.0'; | ||
| 52 | |||
| 53 | # Make it an app command | ||
| 54 | use MooseX::App::Command; | ||
| 55 | extends 'CLIPSeqTools::App'; | ||
| 56 | |||
| 57 | |||
| 58 | ####################################################################### | ||
| 59 | ####################### Load External modules ##################### | ||
| 60 | ####################################################################### | ||
| 61 | use Modern::Perl; | ||
| 62 | use autodie; | ||
| 63 | use namespace::autoclean; | ||
| 64 | use List::Util qw(sum max); | ||
| 65 | use List::Util qw(shuffle); | ||
| 66 | |||
| 67 | |||
| 68 | ####################################################################### | ||
| 69 | ####################### Command line options ###################### | ||
| 70 | ####################################################################### | ||
| 71 | option 'nmer_length' => ( | ||
| 72 | is => 'rw', | ||
| 73 | isa => 'Int', | ||
| 74 | default => 6, | ||
| 75 | documentation => 'the length N of the Nmer.', | ||
| 76 | ); | ||
| 77 | |||
| 78 | option 'permutations' => ( | ||
| 79 | is => 'rw', | ||
| 80 | isa => 'Int', | ||
| 81 | default => 100, | ||
| 82 | documentation => 'the number of permutation to be performed. Use more than 100 to get p-values < 0.01.', | ||
| 83 | ); | ||
| 84 | |||
| 85 | option 'subset' => ( | ||
| 86 | is => 'rw', | ||
| 87 | isa => 'Str', | ||
| 88 | default => '100%', | ||
| 89 | documentation => 'run analysis on random subset. Option specifies the number (if integer) or percent (if % is used) of data to be used.', | ||
| 90 | ); | ||
| 91 | |||
| 92 | |||
| 93 | ####################################################################### | ||
| 94 | ########################## Consume Roles ########################## | ||
| 95 | ####################################################################### | ||
| 96 | with | ||
| 97 | "CLIPSeqTools::Role::Option::Library" => { | ||
| 98 | -alias => { validate_args => '_validate_args_for_library' }, | ||
| 99 | -excludes => 'validate_args', | ||
| 100 | }, | ||
| 101 | "CLIPSeqTools::Role::Option::OutputPrefix" => { | ||
| 102 | -alias => { validate_args => '_validate_args_for_output_prefix' }, | ||
| 103 | -excludes => 'validate_args', | ||
| 104 | }; | ||
| 105 | |||
| 106 | |||
| 107 | ####################################################################### | ||
| 108 | ######################## Interface Methods ######################## | ||
| 109 | ####################################################################### | ||
| 110 | sub validate_args { | ||
| 111 | my ($self) = @_; | ||
| 112 | |||
| 113 | $self->_validate_args_for_library; | ||
| 114 | $self->_validate_args_for_output_prefix; | ||
| 115 | } | ||
| 116 | |||
| 117 | sub run { | ||
| 118 | my ($self) = @_; | ||
| 119 | |||
| 120 | warn "Starting analysis: nmer_enrichment_over_shuffled\n"; | ||
| 121 | |||
| 122 | warn "Validating arguments\n" if $self->verbose; | ||
| 123 | $self->validate_args(); | ||
| 124 | |||
| 125 | warn "Creating reads collection\n" if $self->verbose; | ||
| 126 | my $reads_collection = $self->reads_collection; | ||
| 127 | |||
| 128 | my $subset_factor = 1; | ||
| 129 | if ($self->subset ne '100%') { | ||
| 130 | warn "Calculating subset parameters\n" if $self->verbose; | ||
| 131 | my $records_count = $reads_collection->records_count; | ||
| 132 | if ($self->subset =~ /(.+)%/) { | ||
| 133 | $subset_factor = $1 / 100; | ||
| 134 | } | ||
| 135 | else { | ||
| 136 | $subset_factor = $self->subset / $records_count; | ||
| 137 | } | ||
| 138 | warn "Program will run on approximatelly ".sprintf('%.0f', $subset_factor*100)."% of library\n" if $self->verbose; | ||
| 139 | } | ||
| 140 | |||
| 141 | warn "Counting Nmer occurrences in original and shuffled reads\n" if $self->verbose; | ||
| 142 | my %nmer_stats; | ||
| 143 | $reads_collection->foreach_record_do(sub{ | ||
| 144 | my ($record) = @_; | ||
| 145 | |||
| 146 | return 0 if rand > $subset_factor; | ||
| 147 | |||
| 148 | my $seq = $record->sequence; | ||
| 149 | for my $i (0..length($seq)-$self->nmer_length) { | ||
| 150 | my $nmer = substr($seq, $i, $self->nmer_length); | ||
| 151 | $nmer_stats{$nmer}->{'count'} += $record->copy_number; | ||
| 152 | $nmer_stats{$nmer}->{'collapsed_count'} += 1; | ||
| 153 | } | ||
| 154 | |||
| 155 | my @seq_array = split(//, $seq); | ||
| 156 | for (my $perm = 0; $perm < $self->permutations; $perm++) { | ||
| 157 | my @random_seq_array = shuffle(@seq_array); | ||
| 158 | for (my $i=0; $i < @seq_array - $self->nmer_length + 1; $i++) { | ||
| 159 | my $nmer = join('', @random_seq_array[$i..$i+$self->nmer_length-1]); | ||
| 160 | $nmer_stats{$nmer}->{'sh_count'}->[$perm] += $record->copy_number; | ||
| 161 | $nmer_stats{$nmer}->{'sh_collapsed_count'}->[$perm] += 1; | ||
| 162 | } | ||
| 163 | } | ||
| 164 | |||
| 165 | return 0; | ||
| 166 | }); | ||
| 167 | |||
| 168 | warn "Creating output path\n" if $self->verbose; | ||
| 169 | $self->make_path_for_output_prefix(); | ||
| 170 | |||
| 171 | warn "Calculating p-values and printing results.\n" if $self->verbose; | ||
| 172 | open (my $OUT, '>', $self->o_prefix.'nmer_enrichment_over_shuffled.tab'); | ||
| 173 | say $OUT join("\t", 'nmer', 'count', 'collapsed_count', 'average_sh_count', 'average_sh_collapsed_count', 'enrichment', 'collapsed_enrichment', 'pvalue', 'collapsed_pvalue'); | ||
| 174 | foreach my $nmer (keys %nmer_stats) { | ||
| 175 | next if $nmer =~ /N/; | ||
| 176 | |||
| 177 | my $count = $nmer_stats{$nmer}->{'count'} || 0; | ||
| 178 | my $collapsed_count = $nmer_stats{$nmer}->{'collapsed_count'} || 0; | ||
| 179 | my @sh_counts = @{$nmer_stats{$nmer}->{'sh_count'}}; | ||
| 180 | my @sh_collapsed_counts = @{$nmer_stats{$nmer}->{'sh_collapsed_count'}}; | ||
| 181 | |||
| 182 | map{$_ = 0 if !$_} @sh_counts; # make any undefined values 0 | ||
| 183 | map{$_ = 0 if !$_} @sh_collapsed_counts; # make any undefined values 0 | ||
| 184 | |||
| 185 | my $avg_sh_count = sum(@sh_counts) / $self->permutations || 1; | ||
| 186 | my $avg_collapsed_count = sum(@sh_collapsed_counts) / $self->permutations || 1; | ||
| 187 | |||
| 188 | my $enrichment = $count / $avg_sh_count; | ||
| 189 | my $collapsed_enrichment = $collapsed_count / $avg_collapsed_count; | ||
| 190 | |||
| 191 | my $pvalue = _equal_or_more($count, @sh_counts) / $self->permutations; | ||
| 192 | my $collapsed_pvalue = _equal_or_more($collapsed_count, @sh_collapsed_counts) / $self->permutations; | ||
| 193 | |||
| 194 | say $OUT join("\t", $nmer, $count, $collapsed_count, $avg_sh_count, $avg_collapsed_count, $enrichment, $collapsed_enrichment, $pvalue, $collapsed_pvalue); | ||
| 195 | } | ||
| 196 | } | ||
| 197 | |||
| 198 | |||
| 199 | ####################################################################### | ||
| 200 | ######################## Private Functions ######################## | ||
| 201 | ####################################################################### | ||
| 202 | sub _equal_or_more { | ||
| 203 | my $value = shift; | ||
| 204 | |||
| 205 | return scalar(grep {$_ >= $value} @_); | ||
| 206 | } | ||
| 207 | |||
| 208 | |||
| 209 | |||
| 210 | |||
| 211 | |||
| 212 | |||
| 213 | 1; |