From 0ce82a282e2136977f7d403026840d3eed2e8671 Mon Sep 17 00:00:00 2001
From: gmungoc <g.m.carstairs@dundee.ac.uk>
Date: Mon, 4 Dec 2017 09:21:09 +0000
Subject: [PATCH] JAL-2835 small refactor to only extract transcript
 consequence once

---
 src/jalview/io/vcf/VCFLoader.java |   63 +++++++++++++++++++------------------
 1 file changed, 33 insertions(+), 30 deletions(-)

diff --git a/src/jalview/io/vcf/VCFLoader.java b/src/jalview/io/vcf/VCFLoader.java
index 20e3ccd..421cf38 100644
--- a/src/jalview/io/vcf/VCFLoader.java
+++ b/src/jalview/io/vcf/VCFLoader.java
@@ -902,7 +902,24 @@ public class VCFLoader
     sb.append(forwardStrand ? allele : Dna.reverseComplement(allele));
     String alleles = sb.toString(); // e.g. G,A
 
-    String type = getOntologyTerm(seq, variant, altAlleleIndex);
+    /*
+     * pick out the consequence data (if any) that is for the current allele
+     * and feature (transcript) that matches the current sequence
+     */
+    String consequence = getConsequenceForAlleleAndFeature(variant, CSQ_FIELD,
+            altAlleleIndex, csqAlleleFieldIndex,
+            csqAlleleNumberFieldIndex, seq.getName().toLowerCase(),
+            csqFeatureFieldIndex);
+
+    /*
+     * pick out the ontology term for the consequence type
+     */
+    String type = SequenceOntologyI.SEQUENCE_VARIANT;
+    if (consequence != null)
+    {
+      type = getOntologyTerm(seq, variant, altAlleleIndex,
+            consequence);
+    }
 
     float score = getAlleleFrequency(variant, altAlleleIndex);
 
@@ -912,7 +929,7 @@ public class VCFLoader
 
     sf.setValue(Gff3Helper.ALLELES, alleles);
 
-    addAlleleProperties(variant, seq, sf, altAlleleIndex);
+    addAlleleProperties(variant, seq, sf, altAlleleIndex, consequence);
 
     seq.addSequenceFeature(sf);
 
@@ -931,11 +948,12 @@ public class VCFLoader
    * @param seq
    * @param variant
    * @param altAlleleIndex
+   * @param consequence
    * @return
    * @see http://www.sequenceontology.org/browser/current_svn/term/SO:0001060
    */
   String getOntologyTerm(SequenceI seq, VariantContext variant,
-          int altAlleleIndex)
+          int altAlleleIndex, String consequence)
   {
     String type = SequenceOntologyI.SEQUENCE_VARIANT;
 
@@ -951,10 +969,6 @@ public class VCFLoader
      * can we associate Consequence data with this allele and feature (transcript)?
      * if so, prefer the consequence term from that data
      */
-    String consequence = getConsequenceForAlleleAndFeature(variant,
-            CSQ_FIELD,
-            altAlleleIndex, csqAlleleFieldIndex, csqAlleleNumberFieldIndex,
-            seq.getName().toLowerCase(), csqFeatureFieldIndex);
     if (consequence != null)
     {
       String[] csqFields = consequence.split(PIPE_REGEX);
@@ -1089,9 +1103,12 @@ public class VCFLoader
    * @param sf
    * @param altAlelleIndex
    *          (0, 1..)
+   * @param consequence
+   *          if not null, the consequence specific to this sequence (transcript
+   *          feature) and allele
    */
   protected void addAlleleProperties(VariantContext variant, SequenceI seq,
-          SequenceFeature sf, final int altAlelleIndex)
+          SequenceFeature sf, final int altAlelleIndex, String consequence)
   {
     Map<String, Object> atts = variant.getAttributes();
 
@@ -1105,7 +1122,7 @@ public class VCFLoader
        */
       if (CSQ_FIELD.equals(key))
       {
-        addConsequences(variant, seq, sf, altAlelleIndex);
+        addConsequences(variant, seq, sf, consequence);
         continue;
       }
 
@@ -1163,36 +1180,22 @@ public class VCFLoader
 
   /**
    * Inspects CSQ data blocks (consequences) and adds attributes on the sequence
-   * feature for the current allele (and transcript if applicable)
-   * <p>
-   * Allele matching: if field ALLELE_NUM is present, it must match
-   * altAlleleIndex. If not present, then field Allele value must match the VCF
-   * Allele.
+   * feature.
    * <p>
-   * Transcript matching: if sequence name can be identified to at least one of
-   * the consequences' Feature values, then select only consequences that match
-   * the value (i.e. consequences for the current transcript sequence). If not,
-   * take all consequences (this is the case when adding features to the gene
-   * sequence).
+   * If <code>myConsequence</code> is not null, then this is the specific
+   * consequence data (pipe-delimited fields) that is for the current allele and
+   * transcript (sequence) being processed)
    * 
    * @param variant
    * @param seq
    * @param sf
-   * @param altAlleleIndex
-   *          (0, 1..)
+   * @param myConsequence
    */
   protected void addConsequences(VariantContext variant, SequenceI seq,
-          SequenceFeature sf, int altAlleleIndex)
+          SequenceFeature sf, String myConsequence)
   {
-    /*
-     * first try to identify the matching consequence
-     */
-    String myConsequence = getConsequenceForAlleleAndFeature(variant,
-            CSQ_FIELD, altAlleleIndex, csqAlleleFieldIndex,
-            csqAlleleNumberFieldIndex, seq.getName().toLowerCase(),
-            csqFeatureFieldIndex);
-
     Object value = variant.getAttribute(CSQ_FIELD);
+    // TODO if CSQ not present, try ANN (for SnpEff consequence data)?
 
     if (value == null || !(value instanceof List<?>))
     {
-- 
1.7.10.2