miga-base 0.2.6.4 → 0.2.6.5

data/scripts/trimmed_reads.bash

@@ -1,58 +1,60 @@
 #!/bin/bash
-# Available variables: $PROJECT, $RUNTYPE, $MIGA, $CORES
+# Available variables: $PROJECT, $RUNTYPE, $MIGA, $CORES, $DATASET
 set -e
 SCRIPT="trimmed_reads"
 echo "MiGA: $MIGA"
 echo "Project: $PROJECT"
+# shellcheck source=scripts/miga.bash
 source "$MIGA/scripts/miga.bash" || exit 1
 cd "$PROJECT/data/02.trimmed_reads"
 
 b=$DATASET
 
 # Initialize
-date "+%Y-%m-%d %H:%M:%S %z" > "$DATASET.start"
+miga date > "$DATASET.start"
 
 # Unzip (if necessary)
-[[ -e ../01.raw_reads/$b.1.fastq.gz && ! -e ../01.raw_reads/$b.1.fastq ]] \
-   && gunzip ../01.raw_reads/$b.1.fastq.gz
-[[ -e ../01.raw_reads/$b.2.fastq.gz && ! -e ../01.raw_reads/$b.2.fastq ]] \
-   && gunzip ../01.raw_reads/$b.2.fastq.gz
+[[ -e "../01.raw_reads/$b.1.fastq.gz" && ! -e "../01.raw_reads/$b.1.fastq" ]] \
+  && gunzip "../01.raw_reads/$b.1.fastq.gz"
+[[ -e "../01.raw_reads/$b.2.fastq.gz" && ! -e "../01.raw_reads/$b.2.fastq" ]] \
+  && gunzip "../01.raw_reads/$b.2.fastq.gz"
 
 # Clean existing files
-exists $b.[12].* && rm $b.[12].*
+exists "$b".[12].* && rm "$b".[12].*
 
 # Tag
-FastQ.tag.rb -i ../01.raw_reads/$b.1.fastq -p "$b-" -s "/1" -o $b.1.fastq
-[[ -e ../01.raw_reads/$b.2.fastq ]] \
-   && FastQ.tag.rb -i ../01.raw_reads/$b.2.fastq -p "$b-" -s "/2" -o $b.2.fastq
+FastQ.tag.rb -i "../01.raw_reads/$b.1.fastq" -p "$b-" -s "/1" -o "$b.1.fastq"
+[[ -e "../01.raw_reads/$b.2.fastq" ]] \
+  && FastQ.tag.rb -i "../01.raw_reads/$b.2.fastq" -p "$b-" -s "/2" \
+      -o "$b.2.fastq"
 
 # Trim
-SolexaQA++ dynamictrim $b.[12].fastq -h 20 -d .
-SolexaQA++ lengthsort $b.[12].fastq.trimmed -l 50 -d .
+SolexaQA++ dynamictrim "$b".[12].fastq -h 20 -d .
+SolexaQA++ lengthsort  "$b".[12].fastq.trimmed -l 50 -d .
 
 # Clean adapters
-if [[ -e $b.2.fastq.trimmed.paired ]] ; then
-   scythe -a $MIGA/utils/adapters.fa $b.1.fastq.trimmed.paired \
-      > $b.1.clipped.all.fastq
-   scythe -a $MIGA/utils/adapters.fa $b.2.fastq.trimmed.paired \
-      > $b.2.clipped.all.fastq
-   SolexaQA++ lengthsort $b.[12].clipped.all.fastq -l 50 -d .
-   rm $b.[12].clipped.all.fastq
-   [[ -e $b.1.clipped.all.fastq.single ]] \
-      && mv $b.1.clipped.all.fastq.single $b.1.clipped.single.fastq
-   [[ -e $b.2.clipped.all.fastq.single ]] \
-      && mv $b.2.clipped.all.fastq.single $b.2.clipped.single.fastq
-   mv $b.1.clipped.all.fastq.paired $b.1.clipped.fastq
-   mv $b.2.clipped.all.fastq.paired $b.2.clipped.fastq
-   rm $b.1.clipped.all.fastq.summary.txt &>/dev/null
+if [[ -e "$b.2.fastq.trimmed.paired" ]] ; then
+  scythe -a "$MIGA/utils/adapters.fa" "$b.1.fastq.trimmed.paired" \
+    > "$b.1.clipped.all.fastq"
+  scythe -a "$MIGA/utils/adapters.fa" "$b.2.fastq.trimmed.paired" \
+    > "$b.2.clipped.all.fastq"
+  SolexaQA++ lengthsort "$b".[12].clipped.all.fastq -l 50 -d .
+  rm "$b".[12].clipped.all.fastq
+  [[ -e "$b".1.clipped.all.fastq.single ]] \
+    && mv "$b.1.clipped.all.fastq.single" "$b.1.clipped.single.fastq"
+  [[ -e "$b".2.clipped.all.fastq.single ]] \
+    && mv "$b.2.clipped.all.fastq.single" "$b.2.clipped.single.fastq"
+  mv "$b.1.clipped.all.fastq.paired" "$b.1.clipped.fastq"
+  mv "$b.2.clipped.all.fastq.paired" "$b.2.clipped.fastq"
+  rm -f "$b.1.clipped.all.fastq.summary.txt"
 else
-   scythe -a $MIGA/utils/adapters.fa $b.1.fastq.trimmed.single \
-      > $b.1.clipped.all.fastq
-   SolexaQA++ lengthsort $b.1.clipped.all.fastq -l 50 -d .
-   mv $b.1.clipped.all.fastq.single $b.1.clipped.fastq
+  scythe -a "$MIGA/utils/adapters.fa" "$b.1.fastq.trimmed.single" \
+    > "$b.1.clipped.all.fastq"
+  SolexaQA++ lengthsort "$b.1.clipped.all.fastq" -l 50 -d .
+  mv "$b.1.clipped.all.fastq.single" "$b.1.clipped.fastq"
 fi
-rm $b.[12].*.discard &>/dev/null
+rm -f "$b".[12].*.discard
 
 # Finalize
-date "+%Y-%m-%d %H:%M:%S %z" > "$DATASET.done"
+miga date > "$DATASET.done"
 miga add_result -P "$PROJECT" -D "$DATASET" -r "$SCRIPT"

data/utils/index_metadata.rb

@@ -11,18 +11,18 @@ db = SQLite3::Database.new(ARGV[1])
 db.execute "create table metadata(" +
   "`name` varchar(256), `field` varchar(256), `value` text)"
 
-def searchable(db, k, v)
+def searchable(db, d, k, v)
   db.execute "insert into metadata values(?,?,?)",
-    k.to_s, " #{v.to_s.downcase.gsub(/[^A-Za-z0-9\-]+/, " ")} "
+    d.name, k.to_s, " #{v.to_s.downcase.gsub(/[^A-Za-z0-9\-]+/, " ")} "
 end
 
 p.each_dataset do |name, d|
   next unless d.is_ref?
-  searchable(db, :name, d.name)
+  searchable(db, d, :name, d.name)
   d.metadata.each do |k, v|
     next if [:created, :updated].include? k
     v = v.sorted_ranks.map{ |r| r[1] }.join(" ") if k==:tax
-    searchable(db, k, v)
+    searchable(db, d, k, v)
   end
 end
 

data/utils/ref-tree.R

@@ -0,0 +1,65 @@
+#!/usr/bin/env Rscript
+#
+# @package MiGA
+# @license Artistic-2.0
+#
+
+#= Load stuff
+argv <- commandArgs(trailingOnly=T)
+suppressPackageStartupMessages(library(ape))
+suppressPackageStartupMessages(library(phytools))
+suppressPackageStartupMessages(library(phangorn))
+suppressPackageStartupMessages(library(enveomics.R))
+
+#= Main function
+ref_tree <- function(ani_file, out_base, q_dataset) {
+  a <- read.table(ani_file, sep="\t", header=TRUE, as.is=TRUE)
+  ani.d <- enve.df2dist(a[,1:3], default.d=0.9, max.sim=100)
+  ani.ph <- midpoint(bionj(ani.d))
+  write.tree(ani.ph, paste(out_base, ".nwk", sep=""))
+  pdf(paste(out_base, ".nwk.pdf", sep=""), 7, 7)
+  plot(ani.ph, cex=1/3, type='fan',
+    tip.color=c('red', 'black')[ifelse(ani.ph$tip.label==q_dataset, 1, 2)])
+  add.scale.bar()
+  dev.off()
+}
+
+# Ancilliary functions
+midpoint <- function(tree){
+  dm = cophenetic(tree)
+  tree = unroot(tree)
+  rn = max(tree$edge)+1
+  maxdm = max(dm)
+  ind = which(dm==maxdm,arr=TRUE)[1,]
+  tmproot = Ancestors(tree, ind[1], "parent")
+  tree = phangorn:::reroot(tree, tmproot)
+  edge = tree$edge
+  el = tree$edge.length
+  children = tree$edge[,2]
+  left = match(ind[1], children)
+  tmp = Ancestors(tree, ind[2], "all")
+  tmp= c(ind[2], tmp[-length(tmp)])
+  right = match(tmp, children)
+  if(el[left]>= (maxdm/2)){
+    edge = rbind(edge, c(rn, ind[1]))
+    edge[left,2] = rn
+    el[left] = el[left] - (maxdm/2)
+    el = c(el, maxdm/2)
+  }else{
+    sel = cumsum(el[right])
+    i = which(sel>(maxdm/2))[1]
+    edge = rbind(edge, c(rn, tmp[i]))
+    edge[right[i],2] = rn
+    eltmp = sel[i] - (maxdm/2)
+    el = c(el, el[right[i]] - eltmp)
+    el[right[i]] = eltmp
+  }
+  tree$edge.length = el
+  tree$edge=edge
+  tree$Nnode = tree$Nnode+1
+  phangorn:::reorderPruning(phangorn:::reroot(tree, rn))
+}
+
+#= Main
+ref_tree(ani_file=argv[1], out_base=argv[2], q_dataset=argv[3])
+

data/utils/requirements.txt

@@ -5,7 +5,7 @@ SolexaQA++		SolexaQA++		http://solexaqa.sourceforge.net					Required version: v3
 Scythe			scythe			https://github.com/vsbuffalo/scythe				Required version: 0.991+
 FastQC			fastqc			http://www.bioinformatics.babraham.ac.uk/projects/fastqc
 IDBA			idba_ud			http://i.cs.hku.hk/~alse/hkubrg/projects/idba
-MetaGeneMark		gmhmmp			http://exon.gatech.edu/genemark/license_download.cgi		The folder must contain the key and the scripts
+Prodigal		prodigal		http://prodigal.ornl.gov
 HMMer 3.0+		hmmsearch		http://hmmer.janelia.org/software
 NCBI BLAST+		blastp			ftp://ftp.ncbi.nlm.nih.gov/blast/executables/blast+/LATEST
 R			R			http://www.r-project.org/

metadata

@@ -1,103 +1,103 @@
 --- !ruby/object:Gem::Specification
 name: miga-base
 version: !ruby/object:Gem::Version
-  version: 0.2.6.4
+  version: 0.2.6.5
 platform: ruby
 authors:
 - Luis M. Rodriguez-R
 autorequire: 
 bindir: bin
 cert_chain: []
-date: 2017-04-15 00:00:00.000000000 Z
+date: 2017-05-22 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: rest-client
   requirement: !ruby/object:Gem::Requirement
     requirements:
-    - - ~>
+    - - "~>"
       - !ruby/object:Gem::Version
         version: '1.7'
   type: :runtime
   prerelease: false
   version_requirements: !ruby/object:Gem::Requirement
     requirements:
-    - - ~>
+    - - "~>"
       - !ruby/object:Gem::Version
         version: '1.7'
 - !ruby/object:Gem::Dependency
   name: daemons
   requirement: !ruby/object:Gem::Requirement
     requirements:
-    - - ~>
+    - - "~>"
       - !ruby/object:Gem::Version
         version: '1.2'
-    - - '>='
+    - - ">="
       - !ruby/object:Gem::Version
         version: 1.2.4
   type: :runtime
   prerelease: false
   version_requirements: !ruby/object:Gem::Requirement
     requirements:
-    - - ~>
+    - - "~>"
       - !ruby/object:Gem::Version
         version: '1.2'
-    - - '>='
+    - - ">="
       - !ruby/object:Gem::Version
         version: 1.2.4
 - !ruby/object:Gem::Dependency
   name: json
   requirement: !ruby/object:Gem::Requirement
     requirements:
-    - - ~>
+    - - "~>"
       - !ruby/object:Gem::Version
         version: '1.8'
   type: :runtime
   prerelease: false
   version_requirements: !ruby/object:Gem::Requirement
     requirements:
-    - - ~>
+    - - "~>"
       - !ruby/object:Gem::Version
         version: '1.8'
 - !ruby/object:Gem::Dependency
   name: sqlite3
   requirement: !ruby/object:Gem::Requirement
     requirements:
-    - - ~>
+    - - "~>"
       - !ruby/object:Gem::Version
         version: '1.3'
   type: :runtime
   prerelease: false
   version_requirements: !ruby/object:Gem::Requirement
     requirements:
-    - - ~>
+    - - "~>"
       - !ruby/object:Gem::Version
         version: '1.3'
 - !ruby/object:Gem::Dependency
   name: rake
   requirement: !ruby/object:Gem::Requirement
     requirements:
-    - - ~>
+    - - "~>"
       - !ruby/object:Gem::Version
         version: '11'
   type: :development
   prerelease: false
   version_requirements: !ruby/object:Gem::Requirement
     requirements:
-    - - ~>
+    - - "~>"
       - !ruby/object:Gem::Version
         version: '11'
 - !ruby/object:Gem::Dependency
   name: test-unit
   requirement: !ruby/object:Gem::Requirement
     requirements:
-    - - ~>
+    - - "~>"
       - !ruby/object:Gem::Version
         version: '3'
   type: :development
   prerelease: false
   version_requirements: !ruby/object:Gem::Requirement
     requirements:
-    - - ~>
+    - - "~>"
       - !ruby/object:Gem::Version
         version: '3'
 description: Microbial Genomes Atlas
@@ -108,6 +108,32 @@ extensions: []
 extra_rdoc_files:
 - README.md
 files:
+- Gemfile
+- LICENSE
+- README.md
+- Rakefile
+- actions/add_result.rb
+- actions/add_taxonomy.rb
+- actions/create_dataset.rb
+- actions/create_project.rb
+- actions/daemon.rb
+- actions/date.rb
+- actions/download_dataset.rb
+- actions/find_datasets.rb
+- actions/import_datasets.rb
+- actions/index_taxonomy.rb
+- actions/list_datasets.rb
+- actions/list_files.rb
+- actions/plugins.rb
+- actions/project_info.rb
+- actions/result_stats.rb
+- actions/run_local.rb
+- actions/tax_distributions.rb
+- actions/unlink_dataset.rb
+- bin/miga
+- lib/miga.rb
+- lib/miga/_data/aai-intax.tsv.gz
+- lib/miga/_data/aai-novel.tsv.gz
 - lib/miga/common.rb
 - lib/miga/daemon.rb
 - lib/miga/dataset.rb
@@ -121,18 +147,6 @@ files:
 - lib/miga/tax_index.rb
 - lib/miga/taxonomy.rb
 - lib/miga/version.rb
-- lib/miga.rb
-- test/common_test.rb
-- test/daemon_test.rb
-- test/dataset_test.rb
-- test/metadata_test.rb
-- test/project_test.rb
-- test/remote_dataset_test.rb
-- test/tax_index_test.rb
-- test/taxonomy_test.rb
-- test/test_helper.rb
-- lib/miga/_data/aai-intax.tsv.gz
-- lib/miga/_data/aai-novel.tsv.gz
 - scripts/_distances_functions.bash
 - scripts/_distances_noref_nomulti.bash
 - scripts/_distances_ref_nomulti.bash
@@ -156,38 +170,25 @@ files:
 - scripts/subclades.bash
 - scripts/trimmed_fasta.bash
 - scripts/trimmed_reads.bash
+- test/common_test.rb
+- test/daemon_test.rb
+- test/dataset_test.rb
+- test/metadata_test.rb
+- test/project_test.rb
+- test/remote_dataset_test.rb
+- test/tax_index_test.rb
+- test/taxonomy_test.rb
+- test/test_helper.rb
 - utils/adapters.fa
 - utils/index_metadata.rb
 - utils/mytaxa_scan.R
 - utils/mytaxa_scan.rb
 - utils/plot-taxdist.R
+- utils/ref-tree.R
 - utils/requirements.txt
 - utils/subclades-compile.rb
 - utils/subclades-nj.R
 - utils/subclades.R
-- bin/miga
-- actions/add_result.rb
-- actions/add_taxonomy.rb
-- actions/create_dataset.rb
-- actions/create_project.rb
-- actions/daemon.rb
-- actions/date.rb
-- actions/download_dataset.rb
-- actions/find_datasets.rb
-- actions/import_datasets.rb
-- actions/index_taxonomy.rb
-- actions/list_datasets.rb
-- actions/list_files.rb
-- actions/plugins.rb
-- actions/project_info.rb
-- actions/result_stats.rb
-- actions/run_local.rb
-- actions/tax_distributions.rb
-- actions/unlink_dataset.rb
-- Gemfile
-- Rakefile
-- README.md
-- LICENSE
 homepage: http://enve-omics.ce.gatech.edu/miga
 licenses:
 - Artistic-2.0
@@ -196,25 +197,25 @@ post_install_message:
 rdoc_options:
 - lib
 - README.md
-- --main
+- "--main"
 - README.md
-- --title
+- "--title"
 - MiGA
 require_paths:
 - lib
 required_ruby_version: !ruby/object:Gem::Requirement
   requirements:
-  - - '>='
+  - - ">="
     - !ruby/object:Gem::Version
       version: '1.9'
 required_rubygems_version: !ruby/object:Gem::Requirement
   requirements:
-  - - '>='
+  - - ">="
     - !ruby/object:Gem::Version
       version: '0'
 requirements: []
 rubyforge_project: 
-rubygems_version: 2.0.14
+rubygems_version: 2.6.8
 signing_key: 
 specification_version: 4
 summary: MiGA