| 1 | =head1 NAME | ||
| 2 | |||
| 3 | CLIPSeqTools::PlotApp::cluster_size_and_score_distribution - Create plots for | ||
| 4 | script cluster_size_and_score_distribution. | ||
| 5 | |||
| 6 | =head1 SYNOPSIS | ||
| 7 | |||
| 8 | clipseqtools-plot cluster_size_and_score_distribution [options/parameters] | ||
| 9 | |||
| 10 | =head1 DESCRIPTION | ||
| 11 | |||
| 12 | Create plots for script cluster_size_and_score_distribution. | ||
| 13 | |||
| 14 | =head1 OPTIONS | ||
| 15 | |||
| 16 | Input. | ||
| 17 | --cluster_sizes_file <Str> file with cluster sizes distribution. | ||
| 18 | --cluster_scores_file <Str> file with cluster scores distribution. | ||
| 19 | |||
| 20 | Output | ||
| 21 | --o_prefix <Str> output path prefix. Script will create and | ||
| 22 | add extension to path. Default: ./ | ||
| 23 | |||
| 24 | -v --verbose print progress lines and extra information. | ||
| 25 | -h -? --usage --help print help message | ||
| 26 | |||
| 27 | =cut | ||
| 28 | |||
| 29 | package CLIPSeqTools::PlotApp::cluster_size_and_score_distribution; | ||
| 30 | $CLIPSeqTools::PlotApp::cluster_size_and_score_distribution::VERSION = '1.0.0'; | ||
| 31 | |||
| 32 | # Make it an app command | ||
| 33 | use MooseX::App::Command; | ||
| 34 | extends 'CLIPSeqTools::PlotApp'; | ||
| 35 | |||
| 36 | |||
| 37 | ####################################################################### | ||
| 38 | ####################### Load External modules ##################### | ||
| 39 | ####################################################################### | ||
| 40 | use Modern::Perl; | ||
| 41 | use autodie; | ||
| 42 | use namespace::autoclean; | ||
| 43 | use Statistics::R; | ||
| 44 | |||
| 45 | |||
| 46 | ####################################################################### | ||
| 47 | ####################### Command line options ###################### | ||
| 48 | ####################################################################### | ||
| 49 | option 'cluster_sizes_file' => ( | ||
| 50 | is => 'rw', | ||
| 51 | isa => 'Str', | ||
| 52 | required => 1, | ||
| 53 | documentation => 'file with cluster sizes distribution.', | ||
| 54 | ); | ||
| 55 | |||
| 56 | option 'cluster_scores_file' => ( | ||
| 57 | is => 'rw', | ||
| 58 | isa => 'Str', | ||
| 59 | required => 1, | ||
| 60 | documentation => 'file with cluster scores distribution.', | ||
| 61 | ); | ||
| 62 | |||
| 63 | ####################################################################### | ||
| 64 | ########################## Consume Roles ########################## | ||
| 65 | ####################################################################### | ||
| 66 | with | ||
| 67 | "CLIPSeqTools::Role::Option::OutputPrefix" => { | ||
| 68 | -alias => { validate_args => '_validate_args_for_output_prefix' }, | ||
| 69 | -excludes => 'validate_args', | ||
| 70 | }; | ||
| 71 | |||
| 72 | |||
| 73 | ####################################################################### | ||
| 74 | ######################## Interface Methods ######################## | ||
| 75 | ####################################################################### | ||
| 76 | sub validate_args { | ||
| 77 | my ($self) = @_; | ||
| 78 | |||
| 79 | $self->_validate_args_for_output_prefix; | ||
| 80 | } | ||
| 81 | |||
| 82 | sub run { | ||
| 83 | my ($self) = @_; | ||
| 84 | |||
| 85 | warn "Validating arguments\n" if $self->verbose; | ||
| 86 | $self->validate_args(); | ||
| 87 | |||
| 88 | warn "Creating output path\n" if $self->verbose; | ||
| 89 | $self->make_path_for_output_prefix(); | ||
| 90 | |||
| 91 | warn "Creating plots for cluster sizes with R\n" if $self->verbose; | ||
| 92 | $self->run_R_for_cluster_sizes; | ||
| 93 | |||
| 94 | warn "Creating plots for cluster scores with R\n" if $self->verbose; | ||
| 95 | $self->run_R_for_cluster_scores; | ||
| 96 | } | ||
| 97 | |||
| 98 | sub run_R_for_cluster_scores { | ||
| 99 | my ($self) = @_; | ||
| 100 | |||
| 101 | my $figfile = $self->o_prefix . 'cluster_score_distribution.pdf'; | ||
| 102 | |||
| 103 | # Start R | ||
| 104 | my $R = Statistics::R->new(); | ||
| 105 | |||
| 106 | # Pass arguments to R | ||
| 107 | $R->set('ifile', $self->cluster_scores_file); | ||
| 108 | $R->set('figfile', $figfile); | ||
| 109 | |||
| 110 | # Disable scientific notation | ||
| 111 | $R->run(q{options(scipen=999)}); | ||
| 112 | |||
| 113 | # Read table with data | ||
| 114 | $R->run(q{idata = read.delim(ifile)}); | ||
| 115 | |||
| 116 | # Create column with total number of reads in clusters of each score | ||
| 117 | $R->run(q{idata$contained_reads_per_score = idata$cluster_score * | ||
| 118 | idata$count}); | ||
| 119 | |||
| 120 | # Create groups of scores | ||
| 121 | $R->run(q{mybreaks = c(seq(0,5,1), seq(10,20,5), seq(40,100,20), | ||
| 122 | seq(200,600,200), 1000, Inf)}); | ||
| 123 | $R->run(q{idata$score_group = cut(idata$cluster_score, breaks=mybreaks, | ||
| 124 | dig.lab=4)}); | ||
| 125 | |||
| 126 | # Aggregate (sum) counts for score groups | ||
| 127 | $R->run(q{aggregate_counts = tapply(idata$count, idata$score_group, sum)}); | ||
| 128 | $R->run(q{aggregate_contained_reads_per_score = | ||
| 129 | tapply(idata$contained_reads_per_score, idata$score_group , sum)}); | ||
| 130 | |||
| 131 | # Do plots | ||
| 132 | $R->run(q{pdf(figfile, width=21)}); | ||
| 133 | $R->run(q{par(mfrow = c(1, 3), cex.lab=1.5, cex.axis=1.5, cex.main=1.5, | ||
| 134 | lwd=1.5, oma=c(0, 0, 2, 0), mar=c(9.1, 5.1, 4.1, 2.1))}); | ||
| 135 | |||
| 136 | $R->run(q{plot(aggregate_counts, type="b", xaxt="n", pch=19, xlab = NA, | ||
| 137 | log="y", ylab="Number of clusters", | ||
| 138 | main="Number of clusters with given score")}); | ||
| 139 | $R->run(q{axis(1, at=1:length(aggregate_counts), | ||
| 140 | labels=names(aggregate_counts), las=2)}); | ||
| 141 | $R->run(q{mtext(side = 1, "Cluster score", line = 7, cex=1.2)}); | ||
| 142 | |||
| 143 | $R->run(q{plot((aggregate_counts / sum(aggregate_counts, na.rm=TRUE)) * | ||
| 144 | 100, type="b", xaxt="n", pch=19, xlab = NA, ylim=c(0,100), | ||
| 145 | ylab="Percent of clusters (%)", | ||
| 146 | main="Percent of clusters with given score")}); | ||
| 147 | $R->run(q{axis(1, at=1:length(aggregate_counts), | ||
| 148 | labels=names(aggregate_counts), las=2)}); | ||
| 149 | $R->run(q{mtext(side = 1, "Cluster score", line = 7, cex=1.2)}); | ||
| 150 | |||
| 151 | $R->run(q{plot((aggregate_contained_reads_per_score / | ||
| 152 | sum(aggregate_contained_reads_per_score, na.rm=TRUE)) * 100, type="b", | ||
| 153 | xaxt="n", pch=19, xlab = NA, ylim=c(0,100), | ||
| 154 | ylab="Percent of reads (%)", | ||
| 155 | main="Percent of reads contained in cluster of given score")}); | ||
| 156 | $R->run(q{axis(1, at=1:length(aggregate_counts), | ||
| 157 | labels=names(aggregate_counts), las=2)}); | ||
| 158 | $R->run(q{mtext(side = 1, "Cluster score", line = 7, cex=1.2)}); | ||
| 159 | |||
| 160 | $R->run(q{graphics.off()}); | ||
| 161 | |||
| 162 | # Close R | ||
| 163 | $R->stop(); | ||
| 164 | } | ||
| 165 | |||
| 166 | sub run_R_for_cluster_sizes { | ||
| 167 | my ($self) = @_; | ||
| 168 | |||
| 169 | my $figfile = $self->o_prefix . 'cluster_size_distribution.pdf'; | ||
| 170 | |||
| 171 | # Start R | ||
| 172 | my $R = Statistics::R->new(); | ||
| 173 | |||
| 174 | # Pass arguments to R | ||
| 175 | $R->set('ifile', $self->cluster_sizes_file); | ||
| 176 | $R->set('figfile', $figfile); | ||
| 177 | |||
| 178 | # Disable scientific notation | ||
| 179 | $R->run(q{options(scipen=999)}); | ||
| 180 | |||
| 181 | # Read table with data | ||
| 182 | $R->run(q{idata = read.delim(ifile)}); | ||
| 183 | |||
| 184 | # Create groups of scores | ||
| 185 | $R->run(q{mybreaks = c(10, seq(50,500,50), 1000, Inf)}); | ||
| 186 | $R->run(q{idata$size_group = cut(idata$cluster_size, breaks=mybreaks, | ||
| 187 | dig.lab=4)}); | ||
| 188 | |||
| 189 | # Aggregate (sum) counts for size groups | ||
| 190 | $R->run(q{aggregate_counts = tapply(idata$count, idata$size_group, sum)}); | ||
| 191 | |||
| 192 | # Do plots | ||
| 193 | $R->run(q{pdf(figfile, width=14)}); | ||
| 194 | $R->run(q{par(mfrow = c(1, 2), cex.lab=1.2, cex.axis=1.2, cex.main=1.2, | ||
| 195 | lwd=1.2, oma=c(0, 0, 2, 0), mar=c(8.1, 5.1, 4.1, 2.1))}); | ||
| 196 | |||
| 197 | $R->run(q{plot(aggregate_counts, type="b", xaxt="n", pch=19, xlab = NA, | ||
| 198 | log="y", ylab="Number of clusters", | ||
| 199 | main="Number of clusters with given size")}); | ||
| 200 | $R->run(q{axis(1, at=1:length(aggregate_counts), | ||
| 201 | labels=names(aggregate_counts), las=2)}); | ||
| 202 | $R->run(q{mtext(side = 1, "Cluster size", line = 6, cex=1.2)}); | ||
| 203 | |||
| 204 | $R->run(q{plot((aggregate_counts / sum(aggregate_counts, na.rm=TRUE)) * | ||
| 205 | 100, type="b", xaxt="n", pch=19, xlab = NA, ylim=c(0, 100), | ||
| 206 | ylab="Percent of clusters (%)", | ||
| 207 | main="Percent of clusters with given size")}); | ||
| 208 | $R->run(q{axis(1, at=1:length(aggregate_counts), | ||
| 209 | labels=names(aggregate_counts), las=2)}); | ||
| 210 | $R->run(q{mtext(side = 1, "Cluster size", line = 6, cex=1.2)}); | ||
| 211 | |||
| 212 | $R->run(q{graphics.off()}); | ||
| 213 | |||
| 214 | # Close R | ||
| 215 | $R->stop(); | ||
| 216 | } | ||
| 217 | |||
| 218 | |||
| 219 | 1; |