viral_seq 1.2.7 → 1.4.0

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: 554845dba339d0e06b84c88bc117258516f391bdf58cce015c2669e7b2c6c0d5
-  data.tar.gz: 870280337c90d1f5b9ecbea6e6478d7e2dc22aa70917c6b2ecd94afaa185c1c6
+  metadata.gz: 2bf2afba235cb99f680f10e0913b8e0f715dd2f9c831f2dcba4534a2607685e6
+  data.tar.gz: 8995b0417a1f6ca4de39e26405ab012766a31b1460f5c1c6d11147271f587044
 SHA512:
-  metadata.gz: 54db76e6fd8333ccebb19dee602378ec8dbe5d196ec7bd675e55f65db80cb06ac2ab51ce1f13ab7ea65c0a50ad49978bd3e9581074c497b298f0912858946fa8
-  data.tar.gz: 03d02329192465a9f278715c8a85e3a910e5c5c7252026980d29e669df823a5bdb4be323eeb56f7c9804b71fa8f1763c5a526227f3764315d6eb8e208934ce81
+  metadata.gz: 233485f39d610945794a033c1d2c53680d753ca0284c6b0b9075295352ceb765df11727816dbc061429ccabfb03204c0db82a24a4d1c4a6ebd5a99df770253ff
+  data.tar.gz: ae029b7ae6f530e748ba256a4ba9bb4af95de7e57cbdf49808f1b257794f22830fe3ffbd636a742bd2bae6f5535539b08d0d042b3100ec2b5ef8b59d83098ead

data/Gemfile.lock

@@ -1,12 +1,12 @@
 PATH
   remote: .
   specs:
-    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)
+    viral_seq (1.3.0)
+      colorize (~> 0.1)
+      combine_pdf (~> 1.0, >= 1.0.0)
+      muscle_bio (~> 0.4)
+      prawn (~> 2.3, >= 2.3.0)
+      prawn-table (~> 0.2, >= 0.2.0)
 
 GEM
   remote: https://rubygems.org/

data/README.md

@@ -179,10 +179,28 @@ qc_seqhash.sdrm_hiv_pr(cut_off)
 
 ## Updates
 
+### Version 1.4.0-10132021
+
+  1. Added a function to calculate false detectionr rate (FDR, aka, Benjamini-Hochberg correction) for minority mutations detected in the sequences. `ViralSeq::SeqHash#fdr`
+  2. Updated `bin\tcs_sdrm` script to add FDR value to each DRMs detected.
+
+### Version 1.3.0-08302021
+
+  1. Fixed a bug in the `tcs` pipeline.
+
+### Version 1.2.9-08022021
+
+  1. Fixed a bug when reading the input primer sequences in lowercases.
+  2. Fixed a bug in the method ViralSeq::Math::RandomGaussian
+
+### Version 1.2.8-07292021
+
+  1. Fixed an issue when reading .fastq files containing blank_lines.
+
 ### Version 1.2.7-07152021
 
-  1. Optimzed the workflow of the `tcs` pipeline on raw data with uneven lengths. 
-  `tcs` version to v2.3.5.
+  1. Optimzed the workflow of the `tcs` pipeline on raw data with uneven lengths.
+  `tcs` version to v2.3.6.
 
 
 ### Version 1.2.6-07122021

data/bin/tcs

@@ -152,8 +152,8 @@ begin
     primer[:region] ? region = primer[:region] : region = "region"
     summary_json[:primer_set_name] = region
 
-    cdna_primer = primer[:cdna]
-    forward_primer = primer[:forward]
+    cdna_primer = primer[:cdna].upcase
+    forward_primer = primer[:forward].upcase
 
     export_raw = primer[:export_raw]
     limit_raw = primer[:limit_raw]
@@ -401,7 +401,11 @@ begin
       when 4
         joined_sh = shp.join2(model: :indiv)
       end
-      return joined_sh
+      if joined_sh
+        return joined_sh
+      else
+        joined_sh = ViralSeq::SeqHash.new
+      end
     end
 
     if primer[:end_join]

data/bin/tcs_sdrm

@@ -91,12 +91,12 @@ libs.each do |lib|
   point_mutation_file = File.join(out_lib_dir, (lib_name + "_substitution.csv"))
   point_mutation_out = File.open(point_mutation_file, "w")
   point_mutation_out.puts "region,TCS,AA position,wild type,mutation," +
-                          "number,percentage,95% CI low, 95% CI high, notes"
+                          "number,frequency,95% CI low,95% CI high,fdr,notes"
 
   linkage_file = File.join(out_lib_dir, (lib_name + "_linkage.csv"))
   linkage_out = File.open(linkage_file, "w")
   linkage_out.puts "region,TCS,mutation linkage,number," +
-                   "percentage,95% CI low, 95% CI high, notes"
+                   "frequency,95% CI low, 95% CI high, notes"
 
   aa_report_file = File.join(out_lib_dir, (lib_name + "_aa.csv"))
   aa_report_out = File.open(aa_report_file, "w")
@@ -132,6 +132,7 @@ libs.each do |lib|
       stop_codon_seqs = stop_codon_check[:with_stop_codon]
       filtered_seqs = stop_codon_check[:without_stop_codon]
       poisson_minority_cutoff = filtered_seqs.pm
+      fdr_hash = filtered_seqs.fdr
       summary_hash[:PR] = [
                             seqs.size.to_s,
                             a3g_seqs.size.to_s,
@@ -142,7 +143,7 @@ libs.each do |lib|
       next if filtered_seqs.size < 3
       filtered_seqs.write_nt_fa(File.join(filtered_seq_dir,seq_basename))
 
-      sdrm = filtered_seqs.sdrm_hiv_pr(poisson_minority_cutoff)
+      sdrm = filtered_seqs.sdrm_hiv_pr(poisson_minority_cutoff, fdr_hash)
       point_mutation_list += sdrm[0]
       linkage_list += sdrm[1]
       aa_report_list += sdrm[2]
@@ -155,6 +156,7 @@ libs.each do |lib|
       stop_codon_seqs = stop_codon_check[:with_stop_codon]
       filtered_seqs = stop_codon_check[:without_stop_codon]
       poisson_minority_cutoff = filtered_seqs.pm
+      fdr_hash = filtered_seqs.fdr
       summary_hash[:IN] = [
                             seqs.size.to_s,
                             a3g_seqs.size.to_s,
@@ -165,7 +167,7 @@ libs.each do |lib|
       next if filtered_seqs.size < 3
       filtered_seqs.write_nt_fa(File.join(filtered_seq_dir,seq_basename))
 
-      sdrm = filtered_seqs.sdrm_hiv_in(poisson_minority_cutoff)
+      sdrm = filtered_seqs.sdrm_hiv_in(poisson_minority_cutoff, fdr_hash)
       point_mutation_list += sdrm[0]
       linkage_list += sdrm[1]
       aa_report_list += sdrm[2]
@@ -190,6 +192,7 @@ libs.each do |lib|
       reject_keys = (hypermut_seq_keys | stop_codon_seq_keys)
       filtered_seqs = ViralSeq::SeqHash.new(seqs.dna_hash.reject {|k,v| reject_keys.include?(k) })
       poisson_minority_cutoff = filtered_seqs.pm
+      fdr_hash = filtered_seqs.fdr
       summary_hash[:RT] = [
                             seqs.size.to_s,
                             hypermut_seq_keys.size.to_s,
@@ -200,7 +203,7 @@ libs.each do |lib|
       next if filtered_seqs.size < 3
       filtered_seqs.write_nt_fa(File.join(filtered_seq_dir,seq_basename))
 
-      sdrm = filtered_seqs.sdrm_hiv_rt(poisson_minority_cutoff)
+      sdrm = filtered_seqs.sdrm_hiv_rt(poisson_minority_cutoff, fdr_hash)
       point_mutation_list += sdrm[0]
       linkage_list += sdrm[1]
       aa_report_list += sdrm[2]
@@ -346,7 +349,7 @@ libs.each do |lib|
       title: "Surveillance Drug Resistance Mutations",
       file: point_mutation_file,
       newPDF: "",
-      table_width: [65,55,85,80,60,65,85,85,85,45],
+      table_width: [60,50,70,65,65,60,75,70,70,70,45],
       extra_text: "* Mutation below Poisson cut-off for minority mutations"
     },
     {

data/lib/viral_seq/hivdr.rb

@@ -9,10 +9,13 @@ module ViralSeq
     #   IN codon 53-174 (HXB2 4384-4751)
     # @param cutoff [Integer] cut-off for minimal abundance of a mutation to be called as valid mutation,
     #   can be obtained using ViralSeq::SeqHash#poisson_minority_cutoff function
+    # @param fdr [Hash] hash of events => (false detecton rate)
+    #   can be obtained using ViralSeq::SeqHash#fdr
+    #
     # @return [Array] three elements `[point_mutation_list, linkage_list, report_list]`
     #
     #   # point_mutation_list: two demensional array for the following information,
-    #     # [region,tcs_number,position,wildtype,mutation,count,%,CI_low,CI_high,label]
+    #     # [region,tcs_number,position,wildtype,mutation,count,%,CI_low,CI_high,fdr,label]
     #   # linkage_list: two demensional array for the following information,
     #     # [region,tcs_number,linkage,count,%,CI_low,CI_high,label]
     #   # report_list: two demensional array for the following information,
@@ -20,12 +23,13 @@ module ViralSeq
     # @example identify SDRMs from a FASTA sequence file of HIV PR sequences obtained after MPID-DR sequencing
     #   my_seqhash = ViralSeq::SeqHash.fa('spec/sample_files/sample_dr_sequences/pr.fasta')
     #   p_cut_off = my_seqhash.pm
-    #   pr_sdrm = my_seqhash.sdrm_hiv_pr(p_cut_off)
-    #   puts "region,tcs_number,position,wildtype,mutation,count,%,CI_low,CI_high,label"; pr_sdrm[0].each {|n| puts n.join(',')}
-    #   => region,tcs_number,position,wildtype,mutation,count,%,CI_low,CI_high,label
-    #   => PR,396,30,D,N,247,0.62374,0.57398,0.67163,
-    #   => PR,396,50,I,V,1,0.00253,6.0e-05,0.01399,*
-    #   => PR,396,88,N,D,246,0.62121,0.57141,0.66919,
+    #   fdr_hash = my_seqhash.fdr
+    #   pr_sdrm = my_seqhash.sdrm_hiv_pr(p_cut_off, fdr_hash)
+    #   puts "region,tcs_number,position,wildtype,mutation,count,%,CI_low,CI_high,fdr,label"; pr_sdrm[0].each {|n| puts n.join(',')}
+    #   => region,tcs_number,position,wildtype,mutation,count,%,CI_low,CI_high,fdr,label
+    #   => PR,396,30,D,N,247,0.62374,0.57398,0.67163,0,
+    #   => PR,396,50,I,V,1,0.00253,6.0e-05,0.01399,0.18905,*
+    #   => PR,396,88,N,D,246,0.62121,0.57141,0.66919,0,
     #
     #   puts "region,tcs_number,linkage,count,%,CI_low,CI_high,label"; pr_sdrm[1].each {|n| puts n.join(',')}
     #   => region,tcs_number,linkage,count,%,CI_low,CI_high,label
@@ -136,7 +140,7 @@ module ViralSeq
     #   => PR,98,396,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,99.7475,0.0,0.0,0.0,0.2525,0.0,0.0,0.0,0.0,0.0
     #   => PR,99,396,0.0,0.0,0.0,0.0,100.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0
 
-    def sdrm_hiv_pr(cutoff = 0)
+    def sdrm_hiv_pr(cutoff = 0, fdr_hash = Hash.new(0))
       sequences = self.dna_hash
       region = "PR"
       rf_label = 0
@@ -167,8 +171,9 @@ module ViralSeq
         count_mut_list = mut_list.count_freq
         count_mut_list.each do |m,number|
           ci = ViralSeq::Math::BinomCI.new(number, n_seq)
+          fdr = fdr_hash[number].round(5)
           label = number < cutoff ? "*" : ""
-          point_mutation_list << [region, n_seq, position, wt, m, number, ci.mean.round(5), ci.lower.round(5), ci.upper.round(5), label]
+          point_mutation_list << [region, n_seq, position, wt, m, number, ci.mean.round(5), ci.lower.round(5), ci.upper.round(5), fdr, label]
         end
       end
       point_mutation_list.sort_by! {|record| record[2]}
@@ -229,7 +234,7 @@ module ViralSeq
     # @param (see #sdrm_hiv_pr)
     # @return (see #sdrm_hiv_pr)
 
-    def sdrm_hiv_rt(cutoff = 0)
+    def sdrm_hiv_rt(cutoff = 0, fdr_hash = Hash.new(0))
       sequences = self.dna_hash
       region = "RT"
       rf_label = 1
@@ -280,8 +285,9 @@ module ViralSeq
         count_mut_list = mut_list.count_freq
         count_mut_list.each do |m,number|
           ci = ViralSeq::Math::BinomCI.new(number, n_seq)
+          fdr = fdr_hash[number].round(5)
           label = number < cutoff ? "*" : ""
-          point_mutation_list << ["NRTI", n_seq, position, wt, m, number, ci.mean.round(5), ci.lower.round(5), ci.upper.round(5), label]
+          point_mutation_list << ["NRTI", n_seq, position, wt, m, number, ci.mean.round(5), ci.lower.round(5), ci.upper.round(5), fdr, label]
         end
       end
 
@@ -291,8 +297,9 @@ module ViralSeq
         count_mut_list = mut_list.count_freq
         count_mut_list.each do |m,number|
           ci = ViralSeq::Math::BinomCI.new(number, n_seq)
+          fdr = fdr_hash[number].round(5)
           label = number < cutoff ? "*" : ""
-          point_mutation_list << ["NNRTI", n_seq, position, wt, m, number, ci.mean.round(5), ci.lower.round(5), ci.upper.round(5), label]
+          point_mutation_list << ["NNRTI", n_seq, position, wt, m, number, ci.mean.round(5), ci.lower.round(5), ci.upper.round(5), fdr, label]
         end
       end
 
@@ -365,7 +372,7 @@ module ViralSeq
     # @param (see #sdrm_hiv_pr)
     # @return (see #sdrm_hiv_pr)
 
-    def sdrm_hiv_in(cutoff = 0)
+    def sdrm_hiv_in(cutoff = 0, fdr_hash = Hash.new(0))
       sequences = self.dna_hash
       region = "IN"
       rf_label = 2
@@ -397,8 +404,9 @@ module ViralSeq
         count_mut_list = mut_list.count_freq
         count_mut_list.each do |m,number|
           ci = ViralSeq::Math::BinomCI.new(number, n_seq)
+          fdr = fdr_hash[number].round(5)
           label = number < cutoff ? "*" : ""
-          point_mutation_list << [region, n_seq, position, wt, m, number, ci.mean.round(5), ci.lower.round(5), ci.upper.round(5), label]
+          point_mutation_list << [region, n_seq, position, wt, m, number, ci.mean.round(5), ci.lower.round(5), ci.upper.round(5), fdr, label]
         end
       end
       point_mutation_list.sort_by! {|record| record[2]}

data/lib/viral_seq/math.rb

@@ -31,7 +31,7 @@ module ViralSeq
       def rand
         if (@compute_next_pair = !@compute_next_pair)
           theta = 2 * ::Math::PI * @rng.call
-          scale = @sd * ::Math.sqrt(-2 * Math.log(1 - @rng.call))
+          scale = @sd * ::Math.sqrt(-2 * ::Math.log(1 - @rng.call))
           @g1 = @mean + scale * ::Math.sin(theta)
           @g0 = @mean + scale * ::Math.cos(theta)
         else

data/lib/viral_seq/seq_hash.rb

@@ -116,6 +116,8 @@ module ViralSeq
 
       File.open(fastq_file,'r') do |file|
         file.readlines.collect do |line|
+          line.tr!("\u0000","")
+          next if line == "\n"
           count +=1
           count_m = count % 4
           if count_m == 1
@@ -590,6 +592,37 @@ module ViralSeq
 
     alias_method :pm, :poisson_minority_cutoff
 
+    # calculate false detection rate for minority mutations
+    # Credit: Prof. Michael G. Hudgens from UNC-CH for providing the method for fdr calculation
+    # @param error_rate [Float] estimated sequencing error rate
+    # @return [Hash] pair of mutation frequency to false detection rate. (freq => fdr)
+    # @example calculate FDR for mutations that appeared twice in the sample dataset
+    #   my_seqhash = ViralSeq::SeqHash.fa('spec/sample_files/sample_sequence_for_poisson.fasta')
+    #   fdr_hash = my_seqhash.fdr
+    #   fdr_hash[2].round(5)
+    #   => 0.00726 # means that mutations appear twice have 0.007261748 chance to be caused by residual errors.
+
+    def fdr(error_rate = 0.0001)
+      sequences = self.dna_hash.values
+      if sequences.size == 0
+        return {}
+      else
+        seq_count = self.size
+        observed_hash = variant_for_poisson(sequences)
+        p_unadjusted = []
+        observed_hash.each do |k, v|
+          p_value = 1 - `Rscript -e "cat(pbinom(#{k}-1, #{seq_count}, #{error_rate}))"`.to_f # compute unadjusted exact p-value, ie under null, probability of observing observed_hash[k] or more extreme
+          p_unadjusted += Array.new(v, p_value)
+        end
+        p_fdr = `Rscript -e "cat(p.adjust(c(#{p_unadjusted.join(',')}), 'fdr'))"`.split("\s").count_freq.to_a # controls fdr. aka Benjamini-Hochberg correction
+        vars_pair = observed_hash.to_a
+        fdr_hash = Hash.new(0)
+        (0..(p_fdr.size - 1)).each do |i|
+          fdr_hash[vars_pair[i][0]] = p_fdr[i][0].to_f
+        end
+        return fdr_hash
+      end
+    end #end of #fdr
 
     # align the @dna_hash sequences, return a new ViralSeq::SeqHash object with aligned @dna_hash using MUSCLE
     # @param path_to_muscle [String], path to MUSCLE excutable. if not provided (as default), it will use RubyGem::MuscleBio

data/lib/viral_seq/tcs_core.rb

@@ -305,7 +305,8 @@ module ViralSeq
       end
 
       def general_filter(seq)
-        if seq.size < ($platform_sequencing_length - 1)
+        return false unless seq
+        if seq.size < ($platform_sequencing_length - 10)
           return false
         elsif seq[1..-2] =~ /N/ # sequences with ambiguities except the 1st and last position removed
           return false

data/lib/viral_seq/version.rb

@@ -2,6 +2,6 @@
 # version info and histroy
 
 module ViralSeq
-  VERSION = "1.2.7"
-  TCS_VERSION = "2.3.6"
+  VERSION = "1.4.0"
+  TCS_VERSION = "2.3.8"
 end

metadata

@@ -1,7 +1,7 @@
 --- !ruby/object:Gem::Specification
 name: viral_seq
 version: !ruby/object:Gem::Version
-  version: 1.2.7
+  version: 1.4.0
 platform: ruby
 authors:
 - Shuntai Zhou
@@ -9,7 +9,7 @@ authors:
 autorequire:
 bindir: bin
 cert_chain: []
-date: 2021-07-15 00:00:00.000000000 Z
+date: 2021-10-14 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: bundler