miga-base 0.3.0.0 → 0.3.0.1

data/utils/enveomics/Scripts/BlastTab.metaxaPrep.pl

@@ -0,0 +1,104 @@
+#!/usr/bin/env perl
+
+#
+# @author Luis M. Rodriguez-R <lmrodriguezr at gmail dot com>
+# @update Mar-23-2015
+# @license artistic license 2.0
+#
+
+use warnings;
+use strict;
+use Getopt::Std;
+
+
+sub HELP_MESSAGE { die "
+Usage:
+   $0 [options] genes.txt blast.txt ... > blast_metaxa.txt
+
+   genes.gff2		File containing the genes in any supported format
+   			(see option -f).
+   blast.txt ...	One or more tabular BLAST files.
+   blast_metaxa.txt	Input file for MeTaxa.
+
+   Options:
+   -l <float>		Minimum fraction of the gene aligned to consider a
+   			hit.  By default: 0.75.  Ignored if -f 'no'.
+   -f <str>		Format of the genes prediction.  Any of:
+   			 o gff2: GFF v2 as produced by MetaGeneMark.hmm.
+			 o gff3: GFF v3 with id field in the last column.
+			 o tab: Tabular file with columns gene, gene length,
+			   and contig.
+			 o no: Ignores genes file.
+			By default: gff2.
+   -q			Run quietly.
+   -h			Display this message and exit.
+
+";}
+
+my %o;
+getopts('l:f:qh',\%o);
+my($gff, @blasts) = @ARGV;
+($gff and $#blasts>=0) or &HELP_MESSAGE;
+$o{h} and &HELP_MESSAGE;
+$o{f} ||= "gff2";
+$o{f} = lc $o{f};
+$o{l} ||= 0.75;
+
+my %gene;
+if($o{f} ne 'no'){
+   print STDERR "Reading genes collection.\n" unless $o{q};
+   open GFF, "<", $gff or die "Cannot read file: $gff: $!\n";
+   while(<GFF>){
+      next if /^#/;
+      next if /^\s*$/;
+      chomp;
+      my @ln = split /\t/;
+      if($o{f} eq 'gff2'){
+	 exists $ln[8] or die "Cannot parse line $.: $_\n";
+	 my $id = $ln[8];
+	 $id =~ s/gene_id /gene_id_/;
+	 $ln[0] =~ s/ .*//;
+	 $gene{$id} = [$ln[0], (1+$ln[4]-$ln[3])/3];
+      }elsif($o{f} eq 'gff3'){
+	 exists $ln[8] or die "Cannot parse line $.: $_\n";
+	 $ln[8] =~ /id=([^;]+)/ or die "Cannot parse line $.: $_\n";
+	 my $id = $1;
+	 $ln[0] =~ s/ .*//;
+	 $gene{$id} = [$ln[0], (1+$ln[4]-$ln[3])/3];
+      }elsif($o{f} eq 'tab'){
+         exists $ln[2] or die "Cannot parse line $.: $_\n";
+	 $ln[1]+0 or die "$ln[0]: Length zero.\n";
+         $gene{$ln[0]} = [$ln[2], $ln[1]/3];
+      }else{
+         die "Unsupported format: ".$o{f}.".\n";
+      }
+   }
+   close GFF;
+}
+
+my $i=0;
+my $p=0;
+print STDERR "Generating MeTaxa input.\n" unless $o{q};
+for my $blast (@blasts){
+   print STDERR "  o $blast\n" unless $o{q};
+   open BLAST, "<", $blast or die "Cannot read file: $blast: $!\n";
+   while(<BLAST>){
+      chomp;
+      my @l = split /\t/;
+      $i++;
+      my $ctg;
+      if($o{f} eq 'no'){
+         $ctg = $l[0];
+      }else{
+	 exists $gene{$l[0]} or die "Cannot find contig for gene $l[0].\n";
+	 next unless $l[3] >= $o{l}*$gene{$l[0]}->[1];
+	 $ctg = $gene{$l[0]}->[0];
+      }
+      $l[1] =~ m/gi\|(\d+)\|/ or die "Cannot parse GI in $l[1].\n";
+      print "".join("\t", @l, $ctg, $l[0], $1)."\n";
+      $p++;
+   }
+   close BLAST;
+}
+print STDERR " Found $i results, reported $p.\n" unless $o{q};
+

data/utils/enveomics/Scripts/BlastTab.pairedHits.rb

@@ -0,0 +1,157 @@
+#!/usr/bin/env ruby
+
+#
+# @author Luis M. Rodriguez-R <lmrodriguezr at gmail dot com>
+# @update: Jul-29-2015
+# @license artistic license 2.0
+#
+
+require 'optparse'
+
+opts = {:minscore=>0, :besthits=>0, :orient=>0, :sisprefix=>"_"}
+ARGV << '-h' if ARGV.size==0
+OptionParser.new do |opt|
+   opt.separator "Identifies the best hits of paired-reads."
+   opt.separator ""
+   opt.on("-i", "--blast FILE", "Input BLAST file."){ |v| opts[:blast]=v }
+   opt.on("-s", "--minscore FLOAT", "Minimum (summed) Bit-Score to consider a pair-match."){ |v| opts[:minscore] = v.to_f }
+   opt.on("-b", "--besthits INT", "Outputs top best-hits only (use 0 to output all the paired hits)."){ |v| opts[:besthits]=v.to_i }
+   opt.on("-o", "--orient INT", "Checks the orientation of the hit.  Values are: 0, no checking; 1, same direction; 2,",
+   		"inwards; 3, outwards; 4, different direction (i.e., 2 or 3)."){ |v| opts[:orient]=v.to_i }
+   opt.on("-p", "--sisprefix STR", "Sister read number prefix in the name of the reads.  Escape characters as dots (\\.),",
+   		"parenthesis (\\(, \\), \\[, \\]), or other characters with special meaning in regular expressions",
+		"(\\*, \\+, \\^, \\$, \\|).  This prefix allows regular expressions (for example, use ':|\\.' to use any of",
+		"colon or dot).  Notice that the prefix will not be included in the base name reported in the output."){ |v| opts[:sisprefix]=v }
+   opt.on("-h","--help","Display this screen") do
+      puts opt
+      exit
+   end
+   opt.separator ""
+   opt.separator "Output:"
+   opt.separator "   Tab-delimited flat file, with the following columns:"
+   opt.separator "   1. Query ID (without the \"sister\" identifier)."
+   opt.separator "   2. Subject ID."
+   opt.separator "   3. Bit score (summed from both sister reads)."
+   opt.separator "   4/5. From/To (subject) coordinates for read 1."
+   opt.separator "   6/7. From/To (subject) coordinates for read 2."
+   opt.separator "   8. Reads orientation (1: same direction, 2: inwards, 3: outwards)."
+   opt.separator "   9. Estimated insert size."
+   opt.separator ""
+   opt.separator "Important note: This script assumes that paired hits are next to each other."
+   opt.separator "   If this is not the case (e.g., because the blast was concatenated),"
+   opt.separator "   you must sort the input before running this script."
+   opt.separator ""
+end.parse!
+abort "-i/--blast is mandatory." if opts[:blast].nil?
+abort "-i/--blast must exist." unless File.exists? opts[:blast]
+
+class SingleHit
+   attr_reader :sbj, :score, :orient, :sfrom, :sto, :qfrom, :qto
+   def initialize(blast_ln)
+      blast_ln.chomp!
+      ln = blast_ln.split("\t")
+      @sbj    = ln[1]
+      @score  = ln[11].to_f
+      @qfrom  = ln[6].to_i
+      @qto    = ln[7].to_i
+      @sfrom  = ln[8].to_i
+      @sto    = ln[9].to_i
+      @orient = @sfrom < @sto ? 1 : -1;
+   end
+end
+class DoubleHit
+   attr_reader :name, :sbj, :score, :orient, :hitA, :hitB
+   def initialize(name, hitA, hitB)
+      raise "Trying to set DoubleHit from hits with different subjects" unless hitA.sbj == hitB.sbj
+      @name = name
+      @hitA = hitA
+      @hitB = hitB
+      @sbj = hitA.sbj
+      @score = hitA.score + hitB.score
+      @orient = (hitA.orient == hitB.orient ? 1:
+      		((hitA.orient>0 and hitB.orient<0) ? 2: 3))
+   end
+   def to_s
+      coords = [@hitA.sfrom, @hitB.sfrom, @hitA.sto, @hitB.sto]
+      @name + "\t" + @sbj + "\t" + @score.to_s + "\t" +
+      	@hitA.sfrom.to_s + "\t" + @hitA.sto.to_s + "\t" +
+      	@hitB.sfrom.to_s + "\t" + @hitB.sto.to_s + "\t" +
+	@orient.to_s + "\t" + (coords.max-coords.min).to_s + "\n"
+   end
+end
+class PairedHits
+   attr_reader :name, :hitsA, :hitsB
+   @@minscore = 0
+   @@orient = 0
+   @@besthits = 0
+   def initialize(name)
+      @name = name
+      @hitsA = []
+      @hitsB = []
+      @hits  = []
+   end
+   def hits
+      @hits = []
+      # Search for paired hits
+      @hitsA.each do |hitA|
+         @hitsB.each do |hitB|
+	    if hitA.sbj == hitB.sbj
+	       hit = DoubleHit.new(@name, hitA, hitB)
+	       next if hit.score <= @@minscore # Minimum bit-score check
+	       next if ((1 .. 3).include?(@@orient) and @@orient != hit.orient) # "typical" orientation check
+	       next if (@@orient == 4 and not((2 .. 3).include?(hit.orient))) # "different-orientation" check
+	       @hits.push(hit)
+	    end
+	 end
+      end
+      # Sort the hits
+      @hits.sort! {|x,y| x.score <=> y.score }
+      if @@besthits==0
+         @hits
+      else
+         @hits.take(@@besthits)
+      end
+   end
+   def hitsX(x)
+      if x == 1
+         @hitsA
+      else
+         @hitsB
+      end
+   end
+   # Class methods
+   def PairedHits.minscore=(value)
+      @@minscore = value
+   end
+   def PairedHits.orient=(value)
+      @@orient = value
+   end
+   def PairedHits.besthits=(value)
+      @@besthits = value
+   end
+end
+
+PairedHits.minscore = opts[:minscore]
+PairedHits.orient   = opts[:orient]
+PairedHits.besthits = opts[:besthits]
+
+begin
+   f = File.open(opts[:blast], "r")
+   currPair = PairedHits.new("  ")
+   while(ln = f.gets)
+      m = /^([^\s]*)(?:#{opts[:sisprefix]})([12])/.match(ln)
+      raise "Impossible to parse read name in line #{$.} using sister prefix '#{opts[:sisprefix]}':\n#{ln}"  unless m
+      if m[1] != currPair.name
+	 currPair.hits.each { |hit| puts hit.to_s }
+	 currPair = PairedHits.new(m[1])
+      end
+      currPair.hitsX(m[2].to_i).push(SingleHit.new(ln));
+   end
+   currPair.hits.each { |hit| puts hit.to_s }
+   f.close
+rescue => err
+   $stderr.puts "Exception: #{err}\n\n"
+   err.backtrace.each { |l| $stderr.puts l + "\n" }
+   err
+end
+

data/utils/enveomics/Scripts/BlastTab.recplot2.R

@@ -0,0 +1,40 @@
+#!/usr/bin/env Rscript
+#
+# @author  Luis M. Rodriguez-R
+# @update  Jan-05-2016
+# @license artistic license 2.0
+#
+
+#= Load stuff
+suppressPackageStartupMessages(library(enveomics.R))
+args <- commandArgs(trailingOnly = F)  
+enveomics_R <- file.path(dirname(
+   sub("^--file=", "", args[grep("^--file=", args)])),
+   "lib", "enveomics.R")
+
+#= Generate interface
+opt <- enve.cliopts(enve.recplot2,
+   file.path(enveomics_R, "man", "enve.recplot2.Rd"),
+   positional_arguments=c(1,4),
+   usage="usage: %prog [options] output.Rdata [output.pdf [width height]]",
+   mandatory=c("prefix"),
+   o_desc=list(pos.breaks="Breaks in the positions histogram.",
+     id.breaks="Breaks in the identity histogram.",
+     id.summary="Function summarizing the identity bins. By default: sum."),
+   p_desc=paste("","Produce recruitment plot objects provided that",
+     "BlastTab.catsbj.pl has been previously executed.", sep="\n\t"),
+   ignore=c("plot"),
+
+   defaults=c(id.metric="identity"))
+
+#= Run it!
+if(length(opt$args)>1){
+   args = as.list(opt$args[-1])
+   for(i in 2:3) if(length(args)>=i) args[[i]] <- as.numeric(args[[i]])
+   do.call("pdf", args)
+}else{
+   opt$options[["plot"]] <- FALSE
+}
+rp <- do.call("enve.recplot2", opt$options)
+save(rp, file=opt$args[1])
+if(length(opt$args)>1) dev.off()

data/utils/enveomics/Scripts/BlastTab.seqdepth.pl

@@ -0,0 +1,86 @@
+#!/usr/bin/env perl
+#
+# @author: Luis M Rodriguez-R <lmrodriguezr at gmail dot com>
+# @license: artistic license 2.0
+# @update: Mar-23-2015
+#
+
+use strict;
+use warnings;
+use List::Util qw/min max sum/;
+
+my $fna = shift @ARGV;
+$fna or die "
+Usage:
+   cat blast1... | $0 genes_or_ctgs.fna > genes_or_ctgs.cov
+
+   blast1...		One or more Tabular BLAST files of reads vs genes (or contigs).
+   genes_or_ctgs.fna	A FastA file containing the genes or the contigs (db).
+   genes_or_ctgs.cov	The output file.
+   
+Output:
+   A tab-delimited file with the following columns:
+   1. Subject ID
+   2. Average sequencing depth
+   3. Median sequencing depth
+   4. Number of mapped reads
+   5. Length of the subject sequence
+
+";
+
+my $size  = {};
+my $gene  = {};
+my $reads = {};
+
+SIZE:{
+   local $/=">";
+   print STDERR "== Reading fasta\n";
+   open FNA, "<", $fna or die "Cannot read the file: $fna: $!\n";
+   my $i=0;
+   while(<FNA>){
+      chomp;
+      my @g = split /\n/, $_, 2;
+      next unless $g[1];
+      #$g[1] =~ s/[^A-Za-z]//g;
+      #$size->{$g[0]} = length $g[1];
+      $g[0] =~ s/\s.*//;
+      $size->{$g[0]} = ( $g[1] =~ tr/[A-Za-z]// );
+      print STDERR " Measuring sequence ".($i).": $g[0]      \r" unless ++$i%500;
+   }
+   close FNA;
+   print STDERR " Found $i sequences".(" "x30)."\n";
+}
+
+MAP:{
+   print STDERR "== Reading mapping\n";
+   my $i=0;
+   while(<>){
+      my @ln = split /\t/;
+      $gene->{$ln[1]} ||= [];
+      for my $pos (min($ln[8], $ln[9]) .. max($ln[8], $ln[9])){ ($gene->{$ln[1]}->[$pos]||=0)++ }
+      ($reads->{$ln[1]} ||= 0)++;
+      print STDERR " Saving hit ".($i).": $ln[1]      \r" unless ++$i%5000;
+   }
+   print STDERR " Found $i hits".(" "x30)."\n";
+}
+
+OUT:{
+   print STDERR "== Creating output\n";
+   my $i=0;
+   for my $g (keys %$gene){
+      $gene->{$g}->[$_] ||= 0 for (0 .. $size->{$g});
+      my @sorted = sort {$a <=> $b} @{$gene->{$g}};
+      die "Cannot find gene in $fna: $g.\n" unless exists $size->{$g};
+      printf "%s\t%.6f\t%d\t%d\t%d\n", $g,
+	   sum(@{$gene->{$g}})/$size->{$g},
+	   $sorted[$#sorted/2],
+	   $reads->{$g},
+	   $size->{$g};
+      delete $gene->{$g};
+      print STDERR " Saving sequence $g:".($i)."\r" unless ++$i%500;
+   }
+   print STDERR " Saved $i sequences".(" "x30)."\n";
+}
+
+print STDERR " done.\n";
+

data/utils/enveomics/Scripts/BlastTab.seqdepth_ZIP.pl

@@ -0,0 +1,119 @@
+#!/usr/bin/env perl
+#
+# @author: Luis M Rodriguez-R <lmrodriguezr at gmail dot com>
+# @license: artistic license 2.0
+# @update: Mar-23-2015
+#
+
+use strict;
+use warnings;
+use List::Util qw/min max sum/;
+
+my $fna = shift @ARGV;
+$fna or die "
+Description:
+   Estimates the average sequencing depth of subject sequences (genes or contigs)
+   assuming a Zero-Inflated Poisson distribution (ZIP) to correct for non-covered
+   positions. It uses the corrected method of moments estimators (CMMEs) as described
+   by Beckett et al [1]. Note that [1] has a mistake in eq. (2.4), that should be:
+      pi-hat-MM = 1 - (X-bar / lambda-hat-MM)
+   
+   Also note that a more elaborated mixture distribution can arise from coverage
+   histograms (e.g., see [2] for an additional correction called 'tail distribution'
+   and mixtures involving negative binomial) so take these results cum grano salis.
+
+Usage:
+   cat blast1... | $0 genes_or_ctgs.fna > genes_or_ctgs.cov
+
+   blast1...		One or more Tabular BLAST files of reads vs genes (or contigs).
+   genes_or_ctgs.fna	A FastA file containing the genes or the contigs (db).
+   genes_or_ctgs.cov	The output file.
+   
+Output:
+   A tab-delimited file with the following columns (the one you want is #2):
+   1. Subject ID
+   2. Estimated average sequencing depth (CMME lambda)
+   3. Zero-inflation (CMME pi)
+   4. Observed average sequencing depth
+   5. Observed median sequencing depth
+   6. Observed median sequencing depth excluding zeroes
+   7. Number of mapped reads
+   8. Length of the subject sequence
+
+References:
+   [1] http://anisette.ucs.louisiana.edu/Academic/Sciences/MATH/stage/stat2012.pdf
+   [2] Lindner et al, Bioinformatics, 2013.
+
+";
+
+my $size  = {};
+my $gene  = {};
+my $reads = {};
+
+SIZE:{
+   local $/=">";
+   print STDERR "== Reading fasta\n";
+   open FNA, "<", $fna or die "Cannot read the file: $fna: $!\n";
+   my $i=0;
+   while(<FNA>){
+      chomp;
+      my @g = split /\n/, $_, 2;
+      next unless $g[1];
+      #$g[1] =~ s/[^A-Za-z]//g;
+      #$size->{$g[0]} = length $g[1];
+      $g[0] =~ s/\s.*//;
+      $size->{$g[0]} = ( $g[1] =~ tr/[A-Za-z]// );
+      print STDERR " Measuring sequence ".($i).": $g[0]      \r" unless ++$i%500;
+   }
+   close FNA;
+   print STDERR " Found $i sequences".(" "x30)."\n";
+}
+
+MAP:{
+   print STDERR "== Reading mapping\n";
+   my $i=0;
+   while(<>){
+      my @ln = split /\t/;
+      $gene->{$ln[1]} ||= [];
+      for my $pos (min($ln[8], $ln[9]) .. max($ln[8], $ln[9])){ ($gene->{$ln[1]}->[$pos]||=0)++ }
+      ($reads->{$ln[1]} ||= 0)++;
+      print STDERR " Saving hit ".($i).": $ln[1]      \r" unless ++$i%5000;
+   }
+   print STDERR " Found $i hits".(" "x30)."\n";
+}
+
+OUT:{
+   print STDERR "== Creating output\n";
+   my $i=0;
+   for my $g (keys %$gene){
+      unless(exists $size->{$g}){
+         warn "Warning: Cannot find gene in $fna: $g.\n";
+	 next;
+      }
+      $gene->{$g}->[$_] ||= 0 for (0 .. $size->{$g});
+      die "Hits out-of-boundaries in gene $g: $#{$gene->{$g}} != $size->{$g}.\n" if $#{$gene->{$g}} != $size->{$g};
+      my @sorted = sort {$a <=> $b} @{$gene->{$g}};
+      my @sorted_nz = grep { $_>0 } @sorted;
+      my $xbar = sum(@{$gene->{$g}})/$size->{$g};
+      my $xsqbar = sum(map { ($_ - $xbar)**2 } @{$gene->{$g}})/($size->{$g}-1); 
+      my $var = $xsqbar - $xbar**2;
+      my $lambdaMM = $xbar + ($var/$xbar) - 1;
+      my $piMM = $lambdaMM==0 ? 0 : 1 - $xbar/$lambdaMM;
+      printf "%s\t%.6f\t%.6f\t%.6f\t%d\t%d\t%d\t%d\n", $g,
+	   ($xbar >= $var ? $xbar : $lambdaMM),
+	   ($xbar >= $var ? 0 : $piMM),
+	   #$lambdaMM,
+	   #$piMM,
+	   sum(@{$gene->{$g}})/$size->{$g},
+	   $sorted[$#sorted/2],
+	   $sorted_nz[$#sorted_nz/2],
+	   $reads->{$g},
+	   $size->{$g};
+      delete $gene->{$g};
+      print STDERR " Saving sequence $g:".($i)."   \r" unless ++$i%500;
+   }
+   print STDERR " Saved $i sequences".(" "x30)."   \n";
+}
+
+print STDERR " done.\n";
+