rbbt-dm 1.1.34 → 1.1.36

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA1:
-  metadata.gz: a601a5fe0a12f90cc0729557ffd15c088c47802b
-  data.tar.gz: bc317becbca8ac96f8042ff8a0e8c6dfc2053c3b
+  metadata.gz: d7f4b80f871a403ef1cdcd0253645bc22196e50f
+  data.tar.gz: 986977cae880eb91dc371ebfbe947d24a1b7bb84
 SHA512:
-  metadata.gz: 50beee5add8d5a092a500686f2935085a861e79b8b64d982b8c099de121585fcdce2a5db91ad388a92760b0e8e4ab48745d567b414d4169c2bd609de395bcb26
-  data.tar.gz: 7fca872f255c284cd54539dab39f4444af98ba539302bc317710af6b35b645a61dc73ff72142a8747a7f747e607cfb700136b8714b200ffca4b98d44d1b8c8eb
+  metadata.gz: c56806c3e903c453eeeacd8cbb5d95e26a7e5961d8c14779c0dcfd667ca18100486ac256a9298949a6ef624153e3c83b4fe923702d38edb62bb62f4525c56c96
+  data.tar.gz: 4ee8e20378f61ba5409c054b23e520b70e2175e858b5e6a8ad16bfd8eb98ec95b11a573ad653cebb022d36c2938c72fadc29652e8a8c4a37cad8c03d34e1a2dd

data/lib/rbbt/matrix/barcode.rb

@@ -1,6 +1,6 @@
 require 'rbbt/util/R'
 
-class Matrix
+class RbbtMatrix
   def barcode(outfile, factor = 2)
 
     FileUtils.mkdir_p File.dirname(outfile) unless outfile.nil? or File.exists? File.dirname(outfile)
@@ -45,12 +45,14 @@ rbbt.GE.barcode.mode(#{ R.ruby2R self.data_file }, #{ R.ruby2R outfile }, #{ R.r
     Misc.sensiblewrite(outfile, dumper.stream)
   end
 
-  def activity_cluster(outfile, factor = 2)
+  def activity_cluster(outfile, clusters = 2)
+
+    clusters = Array === clusters ? clusters : (2..clusters).to_a
 
     FileUtils.mkdir_p File.dirname(outfile) unless outfile.nil? or File.exists? File.dirname(outfile)
     cmd =<<-EOF
 source('#{Rbbt.share.R['barcode.R'].find}')
-rbbt.GE.activity_cluster(#{ R.ruby2R self.data_file }, #{ R.ruby2R outfile }, #{R.ruby2R value_type})
+rbbt.GE.activity_cluster(#{ R.ruby2R self.data_file }, #{ R.ruby2R outfile }, #{R.ruby2R key_field}, #{R.ruby2R clusters})
     EOF
 
     R.run(cmd)

data/lib/rbbt/matrix/differential.rb

@@ -1,6 +1,6 @@
 require 'rbbt/util/R'
 
-class Matrix
+class RbbtMatrix
   def differential(main, contrast, path = nil)
     all_samples = self.samples
     if Array === main and Array === contrast
@@ -17,7 +17,7 @@ class Matrix
 
     Persist.persist(name, :tsv, :persist => true, :file => path,
                     :other => {:main => main_samples, :contrast => contrast_samples}, 
-                    :prefix => "Diff", :dir => Matrix.matrix_dir.differential, :no_load => true) do |file|
+                    :prefix => "Diff", :dir => RbbtMatrix.matrix_dir.differential, :no_load => true) do |file|
 
       raise if file.nil?
 

data/lib/rbbt/matrix/knowledge_base.rb

@@ -11,7 +11,7 @@ class KnowledgeBase
   def matrix(name)
     matrix, options = @matrix_registry[name]
 
-    return matrix if Matrix === matrix
+    return matrix if RbbtMatrix === matrix
 
     Path.setup(matrix) if not Path === matrix and File.exists? matrix
 
@@ -35,7 +35,7 @@ class KnowledgeBase
       value_type = TSV.parse_header(data.find).key_field if data
       value_type ||= "Unknown ID"
 
-      Matrix.new data, labels, value_type, format, organism, identifiers
+      RbbtMatrix.new data, labels, value_type, format, organism, identifiers
     else
     end
   end

data/lib/rbbt/matrix.rb

@@ -1,7 +1,7 @@
 require 'rbbt-util'
 require 'rbbt/sources/organism'
 
-class Matrix
+class RbbtMatrix
 
   class << self
     attr_accessor :matrix_dir
@@ -29,6 +29,18 @@ class Matrix
     @identifiers = identifiers 
   end
 
+  def all_fields
+    @all_fields ||= TSV.parse_header(@data_file).all_fields
+  end
+
+  def fields
+    all_fields[1..-1]
+  end
+
+  def key_field
+    all_fields.first
+  end
+
   def samples
     @samples ||= TSV.parse_header(@data_file).fields
   end
@@ -128,13 +140,27 @@ class Matrix
     [main_samples, contrast_samples]
   end
 
+  def to_average(identifiers = nil)
+    name = data_file =~ /:>/ ? File.basename(data_file) : data_file
+
+    file = Persist.persist(data_file, :tsv, :prefix => "Average", :check => [data_file],  :dir => RbbtMatrix.matrix_dir.values, :no_load => true) do
+
+      data = data_file.tsv(:cast => :to_f, :type => :double)
+
+      data.to_list{|v| v.length > 1 ? Misc.mean(v) : v }
+    end
+    subsets = self.subsets
+    matrix = RbbtMatrix.new file, labels, value_type, key_field, organism
+    matrix.subsets = subsets
+    matrix
+  end
 
   def to_gene(identifiers = nil)
     require 'rbbt/tsv/change_id'
 
     name = data_file =~ /:>/ ? File.basename(data_file) : data_file
 
-    file = Persist.persist(data_file, :tsv, :prefix => "Gene", :dir => Matrix.matrix_dir.values, :no_load => true) do
+    file = Persist.persist(data_file, :tsv, :prefix => "Gene", :check => [data_file], :dir => RbbtMatrix.matrix_dir.values, :no_load => true) do
 
       data = data_file.tsv(:cast => :to_f)
 
@@ -145,7 +171,7 @@ class Matrix
       end
     end
     subsets = self.subsets
-    matrix = Matrix.new file, labels, value_type, "Ensembl Gene ID", organism
+    matrix = RbbtMatrix.new file, labels, value_type, "Ensembl Gene ID", organism
     matrix.subsets = subsets
     matrix
   end
@@ -153,11 +179,11 @@ class Matrix
   def to_barcode_ruby(factor = 2)
     name = data_file =~ /:>/ ? File.basename(data_file) : data_file
 
-    file = Persist.persist(data_file, :tsv, :prefix => "Barcode #{factor}", :dir => Matrix.matrix_dir.barcode, :no_load => true) do |filename|
+    file = Persist.persist(data_file, :tsv, :prefix => "Barcode #{factor}", :check => [data_file],  :dir => RbbtMatrix.matrix_dir.barcode, :no_load => true) do |filename|
       barcode_ruby(filename, factor)
     end
     subsets = self.subsets
-    matrix = Matrix.new file, labels, value_type, "Ensembl Gene ID", organism
+    matrix = RbbtMatrix.new file, labels, value_type, key_field, organism
     matrix.subsets = subsets
     matrix
   end
@@ -165,35 +191,35 @@ class Matrix
   def to_barcode(factor = 2)
     name = data_file =~ /:>/ ? File.basename(data_file) : data_file
 
-    file = Persist.persist(data_file, :tsv, :prefix => "Barcode R #{factor}", :dir => Matrix.matrix_dir.barcode, :no_load => true) do |filename|
+    file = Persist.persist(data_file, :tsv, :prefix => "Barcode R #{factor}", :check => [data_file],  :dir => RbbtMatrix.matrix_dir.barcode, :no_load => true) do |filename|
       barcode(filename, factor)
     end
     subsets = self.subsets
-    matrix = Matrix.new file, labels, value_type, "Ensembl Gene ID", organism
+    matrix = RbbtMatrix.new file, labels, value_type, key_field, organism
     matrix.subsets = subsets
     matrix
   end
 
-  def to_activity(factor = 2)
+  def to_activity(clusters = 2)
     require 'rbbt/tsv/change_id'
 
     name = data_file =~ /:>/ ? File.basename(data_file) : data_file
 
-    file = Persist.persist(data_file, :tsv, :prefix => "Activity #{factor}", :dir => Matrix.matrix_dir.barcode, :no_load => true) do |filename|
-      activity_cluster(filename, factor)
+    file = Persist.persist(data_file, :tsv, :prefix => "Activity #{clusters.inspect}", :check => [data_file],  :dir => RbbtMatrix.matrix_dir.barcode, :no_load => true) do |filename|
+      activity_cluster(filename, clusters)
     end
     subsets = self.subsets
-    matrix = Matrix.new file, labels, value_type, "Ensembl Gene ID", organism
+    matrix = RbbtMatrix.new file, labels, value_type, key_field, organism
     matrix.subsets = subsets
     matrix
   end
 
   def tsv(to_gene=true, identifiers = nil)
-    if to_gene and TSV.parse_header(self.data_file).key_field != "Ensembl Gene ID"
+    if to_gene and key_field != "Ensembl Gene ID"
       file =  self.to_gene(identifiers).data_file
-      file.tsv :persist => true, :persist_dir => Matrix.matrix_dir.persist, :type => :double, :merge => true
+      file.tsv :persist => true, :persist_dir => RbbtMatrix.matrix_dir.persist, :type => :double, :merge => true
     else
-      self.data_file.tsv :persist => true, :persist_dir => Matrix.matrix_dir.persist, :merge => true
+      self.data_file.tsv :persist => true, :persist_dir => RbbtMatrix.matrix_dir.persist, :merge => true
     end
   end
 

data/share/R/barcode.R

@@ -63,16 +63,26 @@ rbbt.GE.barcode.mode <- function(matrix_file, output_file, sd.factor = 2, key.fi
   close(file.barcode)
 }
 
-rbbt.GE.activity_cluster <- function(matrix_file, output_file, key.field = "ID"){
+rbbt.GE.activity_cluster <- function(matrix_file, output_file, key.field = "ID", clusters = c(2,3)){
 
     library(mclust)
 
     data = rbbt.tsv.numeric(matrix_file)
 
-    classes = apply(data,2,function(row){Mclust(row)$classification})
+    classes = apply(data,1,function(row){
+                    row.na = is.na(row)
+                    clust = rep(NA, length(row))
+                    if (sum(row.na) <= length(row) - 5){
+                        clust[!row.na] = densityMclust(row[!row.na], prior=priorControl(), G=clusters)$classification
+                    }
+                    clust
+    })
+
+    classes = data.frame(t(classes))
 
     rownames(classes) <- rownames(data)
-    names(classes) <- c("Cluster")
+    names(classes) <- names(data)
 
+    str(classes)
     rbbt.tsv.write(output_file, classes, key.field)
 }

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: rbbt-dm
 version: !ruby/object:Gem::Version
-  version: 1.1.34
+  version: 1.1.36
 platform: ruby
 authors:
 - Miguel Vazquez
 autorequire: 
 bindir: bin
 cert_chain: []
-date: 2017-02-21 00:00:00.000000000 Z
+date: 2017-06-15 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: rbbt-util