viral_seq 1.2.5 → 1.2.6

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: 5fef5f765c741aa14885673a2f980726956ae341a39926929d10ac0c7b4a6ece
-  data.tar.gz: d1e0cb4b691d4aff657209f5f87f02653421dc1bf9ba6dd4f87e43320ad143d5
+  metadata.gz: aee2536198f5579951f7c5f5f1b97ed27fb0bf5459e6dbfe79137a030951d443
+  data.tar.gz: 9ab9b69adf228d3429e02e5d47674ce0a0d137fa6231d940ef8d6cc9abfa1cdf
 SHA512:
-  metadata.gz: 98b18297e15a5fb0eff8706029dcfdecdc5c39df28b6ae3ea8fe5b7611f63f91276f7ed5e459a478b1af25ec25662804290decc923b1c52f946a184d197807be
-  data.tar.gz: 7e359a05ff783971beced635cb2d913c4a9b418391425eba5ce9344b0137b989ea007d1ed5dc07a677a1775d9c73055aad552e971a89ca2207156db5233b243f
+  metadata.gz: d9b5def3e29f819b6d83902676e80fa2d81531cb44132942244125f6b63110d66389e7f9c61d6e9ea3d80ece57d22d95d5c3948c3f7f71d48835b2714794ffe0
+  data.tar.gz: 44b996d9caf9029f2ef522d9058410cdc3b4ec253ccc7bacca5d9ecf66f487624b0b6d18c960b6c7a0af7bfcd2bc72c8c02df59d3c4c7834ab80d4a9eed2a424

data/README.md

@@ -179,11 +179,17 @@ qc_seqhash.sdrm_hiv_pr(cut_off)
 
 ## Updates
 
+### Version 1.2.6-07122021
+
+  1. Optimized the workflow of the `tcs` pipeline in the "end-join/QC/Trimming" section.
+  `tcs` version to v2.3.5.
+
+
 ### Version 1.2.5-06232021
 
   1. Add error rescue and report in the `tcs` pipeline.
     error messages are stored in the .tcs_error file. `tcs` pipeline updated to v2.3.4.
-  2. Use simple majority for the consensus cut-off in the default setting of the `tcs -dr` pipeline. 
+  2. Use simple majority for the consensus cut-off in the default setting of the `tcs -dr` pipeline.
 
 ### Version 1.2.2-05272021
 

data/bin/tcs

@@ -137,7 +137,7 @@ begin
   end
 
   primers = params[:primer_pairs]
-  if primers.empty?
+  if primers.empty? or primers.nil?
     ViralSeq::TcsCore.log_and_abort log, "No primer information. Script terminated."
   end
 
@@ -415,71 +415,65 @@ begin
         joined_sh_raw = end_join(out_dir_raw, primer[:end_join_option], primer[:overlap])
       end
 
-    else
-      File.open(outfile_log, "w") do |f|
-        f.puts JSON.pretty_generate(summary_json)
-      end
-      next
-    end
-
-    if primer[:TCS_QC]
-      ref_start = primer[:ref_start]
-      ref_end = primer[:ref_end]
-      ref_genome = primer[:ref_genome].to_sym
-      indel = primer[:indel]
-      if ref_start == 0
-        ref_start = 0..(ViralSeq::RefSeq.get(ref_genome).size - 1)
-      end
-      if ref_end == 0
-        ref_end = 0..(ViralSeq::RefSeq.get(ref_genome).size - 1)
-      end
-      if primer[:end_join_option] == 1 and primer[:overlap] == 0
-        r1_sh = ViralSeq::SeqHash.fa(outfile_r1)
-        r2_sh = ViralSeq::SeqHash.fa(outfile_r2)
-        r1_sh = r1_sh.hiv_seq_qc(ref_start, (0..(ViralSeq::RefSeq.get(ref_genome).size - 1)), indel, ref_genome)
-        r2_sh = r2_sh.hiv_seq_qc((0..(ViralSeq::RefSeq.get(ref_genome).size - 1)), ref_end, indel, ref_genome)
-        new_r1_seq = r1_sh.dna_hash.each_with_object({}) {|(k, v), h| h[k[0..-4]] = v}
-        new_r2_seq = r2_sh.dna_hash.each_with_object({}) {|(k, v), h| h[k[0..-4]] = v}
-        joined_seq = {}
-        new_r1_seq.each do |seq_name, seq|
-          next unless seq
-          next unless new_r2_seq[seq_name]
-          joined_seq[seq_name] = seq + new_r2_seq[seq_name]
+      if primer[:TCS_QC]
+        ref_start = primer[:ref_start]
+        ref_end = primer[:ref_end]
+        ref_genome = primer[:ref_genome].to_sym
+        indel = primer[:indel]
+        if ref_start == 0
+          ref_start = 0..(ViralSeq::RefSeq.get(ref_genome).size - 1)
+        end
+        if ref_end == 0
+          ref_end = 0..(ViralSeq::RefSeq.get(ref_genome).size - 1)
         end
-        joined_sh = ViralSeq::SeqHash.new(joined_seq)
-
-        if export_raw
-          r1_sh_raw = ViralSeq::SeqHash.fa(outfile_raw_r1)
-          r2_sh_raw = ViralSeq::SeqHash.fa(outfile_raw_r2)
-          r1_sh_raw = r1_sh_raw.hiv_seq_qc(ref_start, (0..(ViralSeq::RefSeq.get(ref_genome).size - 1)), indel, ref_genome)
-          r2_sh_raw = r2_sh_raw.hiv_seq_qc((0..(ViralSeq::RefSeq.get(ref_genome).size - 1)), ref_end, indel, ref_genome)
-          new_r1_seq_raw = r1_sh_raw.dna_hash.each_with_object({}) {|(k, v), h| h[k[0..-4]] = v}
-          new_r2_seq_raw = r2_sh_raw.dna_hash.each_with_object({}) {|(k, v), h| h[k[0..-4]] = v}
-          joined_seq_raw = {}
-          new_r1_seq_raw.each do |seq_name, seq|
+        if primer[:end_join_option] == 1 and primer[:overlap] == 0
+          r1_sh = ViralSeq::SeqHash.fa(outfile_r1)
+          r2_sh = ViralSeq::SeqHash.fa(outfile_r2)
+          r1_sh = r1_sh.hiv_seq_qc(ref_start, (0..(ViralSeq::RefSeq.get(ref_genome).size - 1)), indel, ref_genome)
+          r2_sh = r2_sh.hiv_seq_qc((0..(ViralSeq::RefSeq.get(ref_genome).size - 1)), ref_end, indel, ref_genome)
+          new_r1_seq = r1_sh.dna_hash.each_with_object({}) {|(k, v), h| h[k[0..-4]] = v}
+          new_r2_seq = r2_sh.dna_hash.each_with_object({}) {|(k, v), h| h[k[0..-4]] = v}
+          joined_seq = {}
+          new_r1_seq.each do |seq_name, seq|
             next unless seq
-            next unless new_r2_seq_raw[seq_name]
-            joined_seq_raw[seq_name] = seq + new_r2_seq_raw[seq_name]
+            next unless new_r2_seq[seq_name]
+            joined_seq[seq_name] = seq + new_r2_seq[seq_name]
           end
-          joined_sh_raw = ViralSeq::SeqHash.new(joined_seq_raw)
-        end
-      else
-        joined_sh = joined_sh.hiv_seq_qc(ref_start, ref_end, indel, ref_genome)
+          joined_sh = ViralSeq::SeqHash.new(joined_seq)
+
+          if export_raw
+            r1_sh_raw = ViralSeq::SeqHash.fa(outfile_raw_r1)
+            r2_sh_raw = ViralSeq::SeqHash.fa(outfile_raw_r2)
+            r1_sh_raw = r1_sh_raw.hiv_seq_qc(ref_start, (0..(ViralSeq::RefSeq.get(ref_genome).size - 1)), indel, ref_genome)
+            r2_sh_raw = r2_sh_raw.hiv_seq_qc((0..(ViralSeq::RefSeq.get(ref_genome).size - 1)), ref_end, indel, ref_genome)
+            new_r1_seq_raw = r1_sh_raw.dna_hash.each_with_object({}) {|(k, v), h| h[k[0..-4]] = v}
+            new_r2_seq_raw = r2_sh_raw.dna_hash.each_with_object({}) {|(k, v), h| h[k[0..-4]] = v}
+            joined_seq_raw = {}
+            new_r1_seq_raw.each do |seq_name, seq|
+              next unless seq
+              next unless new_r2_seq_raw[seq_name]
+              joined_seq_raw[seq_name] = seq + new_r2_seq_raw[seq_name]
+            end
+            joined_sh_raw = ViralSeq::SeqHash.new(joined_seq_raw)
+          end
+        else
+          joined_sh = joined_sh.hiv_seq_qc(ref_start, ref_end, indel, ref_genome)
 
-        if export_raw
-          joined_sh_raw = joined_sh_raw.hiv_seq_qc(ref_start, ref_end, indel, ref_genome)
+          if export_raw
+            joined_sh_raw = joined_sh_raw.hiv_seq_qc(ref_start, ref_end, indel, ref_genome)
+          end
         end
-      end
 
-      log.puts Time.now.to_s + "\t" + "Paired TCS number after QC based on reference genome: " + joined_sh.size.to_s
-      summary_json[:combined_tcs_after_qc] = joined_sh.size
-      if primer[:trim]
-        trim_start = primer[:trim_ref_start]
-        trim_end = primer[:trim_ref_end]
-        trim_ref = primer[:trim_ref].to_sym
-        joined_sh = joined_sh.trim(trim_start, trim_end, trim_ref)
-        if export_raw
-          joined_sh_raw = joined_sh_raw.trim(trim_start, trim_end, trim_ref)
+        log.puts Time.now.to_s + "\t" + "Paired TCS number after QC based on reference genome: " + joined_sh.size.to_s
+        summary_json[:combined_tcs_after_qc] = joined_sh.size
+        if primer[:trim]
+          trim_start = primer[:trim_ref_start]
+          trim_end = primer[:trim_ref_end]
+          trim_ref = primer[:trim_ref].to_sym
+          joined_sh = joined_sh.trim(trim_start, trim_end, trim_ref)
+          if export_raw
+            joined_sh_raw = joined_sh_raw.trim(trim_start, trim_end, trim_ref)
+          end
         end
       end
 
@@ -487,11 +481,13 @@ begin
       if export_raw
         joined_sh_raw.write_nt_fa(File.join(out_dir_raw, "combined.raw.fasta"))
       end
+
     end
 
     File.open(outfile_log, "w") do |f|
       f.puts JSON.pretty_generate(summary_json)
     end
+
   end
 
   unless options[:keep]

data/lib/viral_seq/version.rb

@@ -2,6 +2,6 @@
 # version info and histroy
 
 module ViralSeq
-  VERSION = "1.2.5"
-  TCS_VERSION = "2.3.4"
+  VERSION = "1.2.6"
+  TCS_VERSION = "2.3.5"
 end

metadata

@@ -1,7 +1,7 @@
 --- !ruby/object:Gem::Specification
 name: viral_seq
 version: !ruby/object:Gem::Version
-  version: 1.2.5
+  version: 1.2.6
 platform: ruby
 authors:
 - Shuntai Zhou
@@ -9,7 +9,7 @@ authors:
 autorequire:
 bindir: bin
 cert_chain: []
-date: 2021-06-23 00:00:00.000000000 Z
+date: 2021-07-12 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: bundler