viral_seq 1.1.1 → 1.1.2

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: 7a283f3a09cc5d9807e7622cd1ddf27197919955e85d6472b34fc14b66749c03
-  data.tar.gz: 4f90c5a9c7ea0ec148ba7d45ee88dc441f79da67a97654734194a773499ebb8e
+  metadata.gz: '093a8d1d70e251b0748b7035c829eb512991437ffa78fd67387318412e54acf5'
+  data.tar.gz: 1b9d6f6b2cb2ffa8d9cc588b8df096e7ac3840c694bfb241fcf970b738899328
 SHA512:
-  metadata.gz: 385a94eb93c3d8d9116c16a0d8af56ba714ba6191a454076acf881a036de80d1d598f3fcd1a4de841745ca08a1ad3e8bc028a30db9f96c19f3b217ef4583d652
-  data.tar.gz: 714d035b6f65863746cafb120c9cf6eccb8261f3eac69985bad96e5275351eec71aa3b744ee9b462e2dc3e0e199c2d4112386f6a2d7eef89b5b7824c1ab769be
+  metadata.gz: 3853dbfa3f6604d907ec3d77b8c86ec8d885fedcc854c40ca6822ec72e8b2cfe9413bc188aa722a14e4e4f6c9503eca1b36d7f8e0963a5a997c9f0ca8b54fc86
+  data.tar.gz: e5b056cddcf7b87cc30e52c878879cea82d865ea7fc867535767918c30c699d58d6f426518aad02be49916c49f38d9603b0ab27ca6f3625f7a5102ae86863023

data/Gemfile.lock

@@ -1,16 +1,27 @@
 PATH
   remote: .
   specs:
-    viral_seq (1.0.13)
-      colorize (~> 0.1)
-      muscle_bio (~> 0.4)
+    viral_seq (1.1.1)
+      colorize (>= 0.1)
+      combine_pdf (>= 1.0.0)
+      muscle_bio (>= 0.4)
+      prawn (>= 2.3.0)
+      prawn-table (>= 0.2.0)
 
 GEM
   remote: https://rubygems.org/
   specs:
     colorize (0.8.1)
+    combine_pdf (1.0.21)
+      ruby-rc4 (>= 0.1.5)
     diff-lcs (1.3)
     muscle_bio (0.4.0)
+    pdf-core (0.9.0)
+    prawn (2.4.0)
+      pdf-core (~> 0.9.0)
+      ttfunk (~> 1.7)
+    prawn-table (0.2.2)
+      prawn (>= 1.3.0, < 3.0.0)
     rake (13.0.1)
     rspec (3.8.0)
       rspec-core (~> 3.8.0)
@@ -25,6 +36,8 @@ GEM
       diff-lcs (>= 1.2.0, < 2.0)
       rspec-support (~> 3.8.0)
     rspec-support (3.8.0)
+    ruby-rc4 (0.1.5)
+    ttfunk (1.7.0)
 
 PLATFORMS
   ruby

data/README.md

@@ -137,6 +137,12 @@ qc_seqhash.sdrm_hiv_pr(cut_off)
 
 ## Updates
 
+### Version 1.1.2-04262021
+
+  1. Added function `ViralSeq::DRMs.sdrm_json` to export SDRM as json object.
+  2. Added a random string to the temp file names for `muscle_bio` to avoid issues when running scripts in parallel.
+  3. Added `--keep-original` flag to the `tcs` pipeline. 
+
 ### Version 1.1.1-04012021
 
   1. Added warning when paired_raw_sequence less than 0.1% of total_raw_sequence.

data/bin/tcs

@@ -59,6 +59,10 @@ OptionParser.new do |opts|
     exit
   end
 
+  opts.on("--keep-original", "keep raw sequence files") do
+    options[:keep] = true
+  end
+
   opts.on("-v", "--version", "Version info") do
     puts "tcs version: " + ViralSeq::TCS_VERSION.red.bold
     puts "viral_seq version: " + ViralSeq::VERSION.red.bold
@@ -155,6 +159,7 @@ primers.each do |primer|
   forward_primer = primer[:forward]
 
   export_raw = primer[:export_raw]
+  limit_raw = primer[:limit_raw]
 
   unless cdna_primer
     log.puts Time.now.to_s + "\t" + region + " does not have cDNA primer sequence. #{region} skipped."
@@ -261,7 +266,13 @@ primers.each do |primer|
     raw_r1_f = File.open(outfile_raw_r1, 'w')
     raw_r2_f = File.open(outfile_raw_r2, 'w')
 
-    bio_r1.keys.each do |k|
+    if limit_raw
+      raw_keys = bio_r1.keys.sample(limit_raw.to_i)
+    else
+      raw_keys = bio_r1.keys
+    end
+
+    raw_keys.each do |k|
       raw_r1_f.puts k + "_r1"
       raw_r2_f.puts k + "_r2"
       raw_r1_f.puts bio_r1[k]
@@ -474,9 +485,11 @@ primers.each do |primer|
   end
 end
 
-log.puts Time.now.to_s + "\t" + "Removing raw sequence files..."
-File.unlink(r1_f)
-File.unlink(r2_f)
+unless options[:keep]
+  log.puts Time.now.to_s + "\t" + "Removing raw sequence files..."
+  File.unlink(r1_f)
+  File.unlink(r2_f)
+end
 log.puts Time.now.to_s + "\t" + "TCS pipeline successfuly exercuted."
 log.close
 puts "DONE!"

data/lib/viral_seq.rb

@@ -38,5 +38,8 @@ require_relative "viral_seq/version"
 require_relative "viral_seq/tcs_core"
 require_relative "viral_seq/tcs_json"
 require_relative "viral_seq/tcs_dr"
+require_relative "viral_seq/sdrm"
 
 require "muscle_bio"
+require "json"
+require "securerandom"

data/lib/viral_seq/constant.rb

@@ -1,11 +1,41 @@
 module ViralSeq
 
   # array for all amino acid one letter abbreviations
-
   AMINO_ACID_LIST = ["A", "C", "D", "E", "F", "G", "H", "I", "K", "L", "M", "N", "P", "Q", "R", "S", "T", "V", "W", "Y", "*"]
 
-  SDRM_HIV_PR_LIST = {}
-  SDRM_HIV_RT_LIST = {}
-  SDRM_HIV_IN_LIST = {}
-  
+  # R script for tcs_sdrm script
+
+  R_SCRIPT = 'setwd("PATH_TO_FASTA")
+              library(phangorn)
+              library(ape)
+              library(ggplot2)
+              library(scales)
+              library(ggforce)
+              library(cowplot)
+              library(magrittr)
+              library(gridExtra)
+              pdf("OUTPUT_PDF", onefile=T, width=11, height=8.5)
+              fileNames <- list.files()
+              for (fileName in fileNames) {
+              dna <- read.dna(fileName, format="fasta")
+              class(dna)
+              D<- dist.dna(dna, model="raw")
+              pi <- mean(D)
+              dist20 <- quantile(D, prob=c(0.20))
+              alldist <- data.frame(File=fileName, pi, dist20)
+              write.table(alldist,"OUTPUT_CSV",append=TRUE, sep = ",", row.names = FALSE, col.names=FALSE)
+              D2 <- dist.dna(dna, model="TN93")*100
+              def.par <- par(no.readonly = TRUE)
+              par(mfrow=c(1,2))
+              hist<-hist(D, main=fileName, xlab="% Pairwise Distance", ylab="Frequency", col="gray")
+              abline(v=dist20, col="royalblue",lwd=2)
+              abline(v=pi, col="red", lwd=2)
+              legend(x="topright", c("dist20", "pi"), col = c("royalblue", "red"), lwd = c(2,2), cex=0.5)
+              njtree<-NJ(D2)
+              njtreeplot <- plot(njtree, show.tip.label=F, "unrooted", main=fileName)
+              add.scale.bar(cex=0.7, font=2, col="red")
+              }
+              dev.off()'
+
+
 end

data/lib/viral_seq/hivdr.rb

@@ -1,6 +1,6 @@
 
 module ViralSeq
-  class SDRM
+  class SeqHash
 
     # functions to identify SDRMs from a ViralSeq::SeqHash object at HIV PR region.
     #   works for MPID-DR protocol (dx.doi.org/10.17504/protocols.io.useewbe)

data/lib/viral_seq/muscle.rb

@@ -39,8 +39,9 @@ module ViralSeq
 
     def self.align(ref_seq = "", test_seq = "", path_to_muscle = false)
       temp_dir = Dir.home
-      temp_file = File.join(temp_dir, "_temp_muscle_in")
-      temp_aln = File.join(temp_dir, "_temp_muscle_aln")
+      temp_name = "_"  + SecureRandom.alphanumeric
+      temp_file = File.join(temp_dir, temp_name)
+      temp_aln = File.join(temp_dir, (temp_name + "_aln"))
       name = ">test"
       temp_in = File.open(temp_file,"w")
       temp_in.puts ">ref"

data/lib/viral_seq/sdrm.rb

@@ -1,43 +1,109 @@
 module ViralSeq
   class DRMs
-    def initialize (mutation_list = {})
-      @mutation_list = mutation_list
-    end
-
-    attr_accessor :mutation_list
-  end
+    class << self
 
-  def self.sdrm_hiv_pr(seq_hash)
-  end
-
-  def self.sdrm_hiv_rt(seq_hash)
-  end
+      # function to retrieve sdrm positions as a hash
+      # @param ref_option [Symbol], name of reference genomes, options are `:hiv_pr`, `:hiv_rt`, `:hiv_in`, `hcv_ns5a`
+      # @return [Hash] Hash of :position_number => [ 'wildtype_codon', ['mutation_codons']]
+      def sdrm_hash(options)
+        sdrm = {}
+        case options
+        when :hcv_ns5a
+          sdrm[28] = ['M',['T']]
+          sdrm[30] = ['L',['H','K','R','Q','A','S','D']]
+          sdrm[31] = ['L',['M','V','F']]
+          sdrm[32] = ['P',['L']]
+          sdrm[44] = ['K',['R']]
+          sdrm[58] = ['H',['D','P','S']]
+          sdrm[64] = ['T',['A','S']]
+          sdrm[77] = ['P',['A','S']]
+          sdrm[78] = ['R',['K']]
+          sdrm[79] = ['T',['A']]
+          sdrm[83] = ['T',['M']]
+          sdrm[85] = ['S',['N','H','Y']]
+          sdrm[92] = ['A',['P','T','K','E']]
+          sdrm[93] = ['Y',['C','F','H','N']]
+          sdrm[107] = ['K',['T','S']]
+          sdrm[121] = ['I',['V']]
+          sdrm[135] = ['T',['A']]
+        when :nrti
+          sdrm[41] = ['M',['L']]
+          sdrm[65] = ['K',['R']]
+          sdrm[67] = ['D',['N','G','E']]
+          sdrm[69] = ['T',['D']]
+          sdrm[70] = ['K',['R','E']]
+          sdrm[74] = ['L',['V','I']]
+          sdrm[75] = ['V',['M','T','A','S']]
+          sdrm[77] = ['F',['L']]
+          sdrm[115] = ['Y',['F']]
+          sdrm[116] = ['F',['Y']]
+          sdrm[151] = ['Q',['M']]
+          sdrm[184] = ['M',['V','I']]
+          sdrm[210] = ['L',['W']]
+          sdrm[215] = ["T",["Y","F","I","C","D","V","E"]]
+          sdrm[219] = ["K",["Q","E","N","R"]]
+        when :nnrti
+          sdrm[100] = ['L',['I']]
+          sdrm[101] = ['K',['E','P']]
+          sdrm[103] = ['K',['N','S']]
+          sdrm[106] = ['V',['M','A']]
+          sdrm[179] = ['V',['F','D']]
+          sdrm[181] = ['Y',['C','I','V']]
+          sdrm[188] = ['Y',['L','H','C']]
+          sdrm[190] = ['G',['A','S','E']]
+          sdrm[225] = ['P',['H']]
+          sdrm[230] = ['M',['L']]
+        when :hiv_pr
+          sdrm[23] = ['L',['I']]
+          sdrm[24] = ['L',['I']]
+          sdrm[30] = ['D',['N']]
+          sdrm[32] = ['V',['I']]
+          sdrm[46] = ['M',['I','L']]
+          sdrm[47] = ['I',['V','A']]
+          sdrm[48] = ['G',['V','M']]
+          sdrm[50] = ['I',['V','L']]
+          sdrm[53] = ['F',['L']]
+          sdrm[54] = ['I',['V','L','M','T','A','S']]
+          sdrm[73] = ['G',['S','T','C','A']]
+          sdrm[76] = ['L',['V']]
+          sdrm[82] = ['V',['A','T','S','F','L','C','M']]
+          sdrm[83] = ['N',['D']]
+          sdrm[84] = ['I',['V','A','C']]
+          sdrm[88] = ['N',['D','S']]
+          sdrm[90] = ['L',['M']]
+        when :hiv_in
+          sdrm[66] = ['T',['A','I','K']]
+          sdrm[74] = ['L',['M']]
+          sdrm[92] = ['E',['Q']]
+          sdrm[95] = ['Q',['K']]
+          sdrm[97] = ['T',['A']]
+          sdrm[121] = ['F',['Y']]
+          sdrm[140] = ['G',['A','S','C']]
+          sdrm[143] = ["Y",["C","H","R"]]
+          sdrm[147] = ['S',['G']]
+          sdrm[148] = ['Q',['H','K','R']]
+          sdrm[155] = ['N',['S','H']]
+        else raise "Input option `#{options}` for ViralSeq::Sequence.sdrm not supported"
+        end
+        return sdrm
+      end # end of #sdrm_hash
 
-  def self.sdrm_hiv_in(seq_hash)
-  end
-
-  def self.list_from_json(file)
-  end
-
-  def self.list_from_csv(file)
-  end
-
-  def self.export_list_hiv_pr(file, format = :json)
-    if foramt == :json
+      # function to export SDRM positions as json object
+      # @param (see #sdrm_hash)
+      # @return [String] json String of SDRM positions
 
+      def sdrm_json(options)
+        sdrm = ViralSeq::DRMs.sdrm_hash(options)
+        json_array = []
+        sdrm.each do |pos, muts|
+          mutation = {}
+          mutation[:position] = pos
+          mutation[:wildtypeCodon] = muts[0]
+          mutation[:mutationCodons] = muts[1]
+          json_array << mutation
+        end
+        JSON.pretty_generate(json_array)
+      end
     end
   end
-
-  def self.export_list_hiv_rt(file, format = :json)
-
-  end
-
-  def self.export_list_hiv_in(file, format = :json)
-
-  end
-
-  def drm_analysis(seq_hash)
-    mutation_list = self.mutation_list
-
-  end
 end

data/lib/viral_seq/sequence.rb

@@ -113,7 +113,7 @@ module ViralSeq
     def sdrm(option, start_aa = 1)
       aa_array = self.aa_array
       out_hash = {}
-      sdrm = sdrm_hash(option)
+      sdrm = ViralSeq::DRMs.sdrm_hash(option)
       aa_length = aa_array.size
       end_aa = start_aa + aa_length - 1
       (start_aa..end_aa).each do |position|
@@ -535,88 +535,5 @@ module ViralSeq
       return aa_out
     end # end of #amino_acid_2
 
-    # sdrm position hash
-    def sdrm_hash(options)
-      sdrm = {}
-      case options
-      when :hcv_ns5a
-        sdrm[28] = ['M',['T']]
-        sdrm[30] = ['L',['H','K','R','Q','A','S','D']]
-        sdrm[31] = ['L',['M','V','F']]
-        sdrm[32] = ['P',['L']]
-        sdrm[44] = ['K',['R']]
-        sdrm[58] = ['H',['D','P','S']]
-        sdrm[64] = ['T',['A','S']]
-        sdrm[77] = ['P',['A','S']]
-        sdrm[78] = ['R',['K']]
-        sdrm[79] = ['T',['A']]
-        sdrm[83] = ['T',['M']]
-        sdrm[85] = ['S',['N','H','Y']]
-        sdrm[92] = ['A',['P','T','K','E']]
-        sdrm[93] = ['Y',['C','F','H','N']]
-        sdrm[107] = ['K',['T','S']]
-        sdrm[121] = ['I',['V']]
-        sdrm[135] = ['T',['A']]
-      when :nrti
-        sdrm[41] = ['M',['L']]
-        sdrm[65] = ['K',['R']]
-        sdrm[67] = ['D',['N','G','E']]
-        sdrm[69] = ['T',['D']]
-        sdrm[70] = ['K',['R','E']]
-        sdrm[74] = ['L',['V','I']]
-        sdrm[75] = ['V',['M','T','A','S']]
-        sdrm[77] = ['F',['L']]
-        sdrm[115] = ['Y',['F']]
-        sdrm[116] = ['F',['Y']]
-        sdrm[151] = ['Q',['M']]
-        sdrm[184] = ['M',['V','I']]
-        sdrm[210] = ['L',['W']]
-        sdrm[215] = ["T",["Y","F","I","C","D","V","E"]]
-        sdrm[219] = ["K",["Q","E","N","R"]]
-      when :nnrti
-        sdrm[100] = ['L',['I']]
-        sdrm[101] = ['K',['E','P']]
-        sdrm[103] = ['K',['N','S']]
-        sdrm[106] = ['V',['M','A']]
-        sdrm[179] = ['V',['F','D']]
-        sdrm[181] = ['Y',['C','I','V']]
-        sdrm[188] = ['Y',['L','H','C']]
-        sdrm[190] = ['G',['A','S','E']]
-        sdrm[225] = ['P',['H']]
-        sdrm[230] = ['M',['L']]
-      when :hiv_pr
-        sdrm[23] = ['L',['I']]
-        sdrm[24] = ['L',['I']]
-        sdrm[30] = ['D',['N']]
-        sdrm[32] = ['V',['I']]
-        sdrm[46] = ['M',['I','L']]
-        sdrm[47] = ['I',['V','A']]
-        sdrm[48] = ['G',['V','M']]
-        sdrm[50] = ['I',['V','L']]
-        sdrm[53] = ['F',['L']]
-        sdrm[54] = ['I',['V','L','M','T','A','S']]
-        sdrm[73] = ['G',['S','T','C','A']]
-        sdrm[76] = ['L',['V']]
-        sdrm[82] = ['V',['A','T','S','F','L','C','M']]
-        sdrm[83] = ['N',['D']]
-        sdrm[84] = ['I',['V','A','C']]
-        sdrm[88] = ['N',['D','S']]
-        sdrm[90] = ['L',['M']]
-      when :hiv_in
-        sdrm[66] = ['T',['A','I','K']]
-        sdrm[74] = ['L',['M']]
-        sdrm[92] = ['E',['Q']]
-        sdrm[95] = ['Q',['K']]
-        sdrm[97] = ['T',['A']]
-        sdrm[121] = ['F',['Y']]
-        sdrm[140] = ['G',['A','S','C']]
-        sdrm[143] = ["Y",["C","H","R"]]
-        sdrm[147] = ['S',['G']]
-        sdrm[148] = ['Q',['H','K','R']]
-        sdrm[155] = ['N',['S','H']]
-      else raise "Input option `#{options}` for ViralSeq::Sequence.sdrm not supported"
-      end
-      return sdrm
-    end
   end # end of ViralSeq::Sequence
 end # end of ViralSeq

data/lib/viral_seq/tcs_dr.rb

@@ -16,7 +16,7 @@ module ViralSeq
                   :ref_genome=>"HXB2",
                   :ref_start=>2648,
                   :ref_end=>3257,
-                  :indel=>true,
+                  :indel=>false,
                   :trim=>false},
                  {:region=>"PR",
                   :cdna=>

data/lib/viral_seq/version.rb

@@ -2,6 +2,6 @@
 # version info and histroy
 
 module ViralSeq
-  VERSION = "1.1.1"
-  TCS_VERSION = "2.3.0"
+  VERSION = "1.1.2"
+  TCS_VERSION = "2.3.1"
 end

data/viral_seq.gemspec

@@ -29,11 +29,22 @@ Gem::Specification.new do |spec|
   spec.add_development_dependency "rake", "~> 13.0"
   spec.add_development_dependency "rspec", "~> 3.0"
 
+  # This gem will work with Ruby version 2.5.0 or greater...
+  spec.required_ruby_version = '>= 2.5'
+  # This gem will work with RubyGem version 1.3.6 or greater...
+  spec.required_rubygems_version = '>= 1.3.6'
+
   # muscle_bio gem required
   spec.add_runtime_dependency "muscle_bio", "~> 0.4"
 
   # colorize gem required
   spec.add_runtime_dependency "colorize", "~> 0.1"
 
+  spec.add_runtime_dependency "prawn", "~> 2.3", '>= 2.3.0'
+
+  spec.add_runtime_dependency "prawn-table", "~> 0.2", '>= 0.2.0'
+
+  spec.add_runtime_dependency "combine_pdf", "~> 1.0", '>= 1.0.0'
+
   spec.requirements << 'R required for some functions'
 end

metadata

@@ -1,7 +1,7 @@
 --- !ruby/object:Gem::Specification
 name: viral_seq
 version: !ruby/object:Gem::Version
-  version: 1.1.1
+  version: 1.1.2
 platform: ruby
 authors:
 - Shuntai Zhou
@@ -9,7 +9,7 @@ authors:
 autorequire: 
 bindir: bin
 cert_chain: []
-date: 2021-04-01 00:00:00.000000000 Z
+date: 2021-04-27 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: bundler
@@ -81,6 +81,66 @@ dependencies:
     - - "~>"
       - !ruby/object:Gem::Version
         version: '0.1'
+- !ruby/object:Gem::Dependency
+  name: prawn
+  requirement: !ruby/object:Gem::Requirement
+    requirements:
+    - - "~>"
+      - !ruby/object:Gem::Version
+        version: '2.3'
+    - - ">="
+      - !ruby/object:Gem::Version
+        version: 2.3.0
+  type: :runtime
+  prerelease: false
+  version_requirements: !ruby/object:Gem::Requirement
+    requirements:
+    - - "~>"
+      - !ruby/object:Gem::Version
+        version: '2.3'
+    - - ">="
+      - !ruby/object:Gem::Version
+        version: 2.3.0
+- !ruby/object:Gem::Dependency
+  name: prawn-table
+  requirement: !ruby/object:Gem::Requirement
+    requirements:
+    - - "~>"
+      - !ruby/object:Gem::Version
+        version: '0.2'
+    - - ">="
+      - !ruby/object:Gem::Version
+        version: 0.2.0
+  type: :runtime
+  prerelease: false
+  version_requirements: !ruby/object:Gem::Requirement
+    requirements:
+    - - "~>"
+      - !ruby/object:Gem::Version
+        version: '0.2'
+    - - ">="
+      - !ruby/object:Gem::Version
+        version: 0.2.0
+- !ruby/object:Gem::Dependency
+  name: combine_pdf
+  requirement: !ruby/object:Gem::Requirement
+    requirements:
+    - - "~>"
+      - !ruby/object:Gem::Version
+        version: '1.0'
+    - - ">="
+      - !ruby/object:Gem::Version
+        version: 1.0.0
+  type: :runtime
+  prerelease: false
+  version_requirements: !ruby/object:Gem::Requirement
+    requirements:
+    - - "~>"
+      - !ruby/object:Gem::Version
+        version: '1.0'
+    - - ">="
+      - !ruby/object:Gem::Version
+        version: 1.0.0
 description: |-
   A Ruby Gem with bioinformatics tools for processing viral NGS data.
                             Specifically for Primer-ID sequencing and HIV drug resistance analysis.
@@ -142,12 +202,12 @@ required_ruby_version: !ruby/object:Gem::Requirement
   requirements:
   - - ">="
     - !ruby/object:Gem::Version
-      version: '0'
+      version: '2.5'
 required_rubygems_version: !ruby/object:Gem::Requirement
   requirements:
   - - ">="
     - !ruby/object:Gem::Version
-      version: '0'
+      version: 1.3.6
 requirements:
 - R required for some functions
 rubygems_version: 3.2.2