Merge branch 'hotfix/v2.1.2' into develop

# Conflicts:
#	pom.xml

CHANGELOG

@@ -1,4 +1,11 @@
 
+MiXCR 2.1.2 (20 Apr 2017)
+========================
+
+-- Fixed rear IllegalArgumentException in `extendAlignments` action
+-- Fixed rear NPE in `align` action
+
+
 MiXCR 2.1.1 ( 3 Mar 2017)
 ========================
 

doc/assemble.rst

@@ -184,9 +184,7 @@ can set ``maxBadPointsPercent`` to zero:
 Separation of clones with same CDR3 (clonal sequence) but different V/J/C genes
 -------------------------------------------------------------------------------
 
-Since v1.8 MiXCR by default separates clones with equal clonal sequence and different V and J genes
-and optionally can separate clones with different C genes (e.g. do distinguish clones with different
-IG isotype).
+Since v1.8 MiXCR can separates clones with equal clonal sequence and different V, J and C (e.g. do distinguish clones with different IG isotype) genes.
 
 To make analysis more robust to sequencing errors there is an additional clustering step to shrink
 artificial diversity generated by this separation mechanism.

doc/quickstart.rst

@@ -174,11 +174,11 @@ For the full length cDNA-based immunoglobulin repertoire analysis we generally r
 
   ::
   
-    > mixcr align -p kaligner2 –s hsa -r alignmentReport.txt -OreadsLayout=Collinear \
+    > mixcr align -p kaligner2 -s hsa -r alignmentReport.txt -OreadsLayout=Collinear \
       -OvParameters.geneFeatureToAlign=VTranscript read_R1.fastq.gz read_R2.fastq.gz \
       alignments.vdjca
      
-  Option ``-s`` allows to specify species (e.g. homo sapiens - ``hsa``, mus musculus - ``mmu``). Parameter ``–OreadsLayout`` allow us to set paired-end reads orientation (``Collinear``, ``Opposite``, ``Unknown``). Note, that after MiGEC analysis paired-end read pairs are in ``Collinear`` orientation.
+  Option ``-s`` allows to specify species (e.g. homo sapiens - ``hsa``, mus musculus - ``mmu``). Parameter ``-OreadsLayout`` allow us to set paired-end reads orientation (``Collinear``, ``Opposite``, ``Unknown``). Note, that after MiGEC analysis paired-end read pairs are in ``Collinear`` orientation.
 
   Instead of KAligner2, default MiXCR aligner can be used as well, but it may miss immunoglobulin subvariants that contain several nucleotide-lengths indels within the V gene segment.
 
@@ -187,13 +187,13 @@ For the full length cDNA-based immunoglobulin repertoire analysis we generally r
   ::
   
     > mixcr assemble -p default_affine -r assembleReport.txt -OassemblingFeatures=VDJRegion \
-      –OseparateByC=true -OqualityAggregationType=Average \
+      -OseparateByC=true -OqualityAggregationType=Average \
       -OclusteringFilter.specificMutationProbability=1E-5 -OmaxBadPointsPercent=0 \
       alignments.vdjca clones.clns
 
   ``default_affine`` parameter is specifically required for the data aligned using KAligner2 (use this option only if ``-p kaligner2`` was used on the alignemnt step)
 
-  ``–OseparateByC=true`` separates clones with different antibody isotype.
+  ``-OseparateByC=true`` separates clones with different antibody isotype.
 
   Set ``-OcloneClusteringParameters=null`` parameter to switch off the frequency-based correction of PCR errors.
 
@@ -205,9 +205,9 @@ For the full length cDNA-based immunoglobulin repertoire analysis we generally r
 
   ::
 
-    > mixcr exportClones –c IGH -o -t clones.clns clones.txt
+    > mixcr exportClones -c IGH -o -t clones.clns clones.txt
 
-  where options ``-o`` and ``-t`` filter off the out-of-frame and stop codon containing clonotypes, respectively, and ``–c`` indicates which chain will be extracted (e.g. ``IGH``, ``IGL``).
+  where options ``-o`` and ``-t`` filter off the out-of-frame and stop codon containing clonotypes, respectively, and ``-c`` indicates which chain will be extracted (e.g. ``IGH``, ``IGL``).
 
 
 .. _ref-exampleRnaSeq:

pom.xml

@@ -32,7 +32,7 @@
 
     <groupId>com.milaboratory</groupId>
     <artifactId>mixcr</artifactId>
-    <version>2.1.2-SNAPSHOT</version>
+    <version>2.2-SNAPSHOT</version>
     <packaging>jar</packaging>
     <name>MiXCR</name>
 

src/main/java/com/milaboratory/mixcr/partialassembler/TargetMerger.java

@@ -128,6 +128,42 @@ public HitMappingRecord(VDJCGene gene, Alignment<NucleotideSequence>[] alignment
         }
     }
 
+    static boolean hasAlignments(AlignedTarget target, GeneType geneType) {
+        for (VDJCHit l : target.getAlignments().getHits(geneType))
+            if (l.getAlignment(target.getTargetId()) != null)
+                return true;
+        return false;
+    }
+
+    @SuppressWarnings("unchecked")
+    static List<HitMappingRecord> extractSortedHits(AlignedTarget targetLeft, AlignedTarget targetRight, GeneType geneType) {
+        // Fast calculation for targets from the same PE-read (or multi-read)
+        if (targetLeft.getAlignments() == targetRight.getAlignments()) {
+            VDJCHit[] hits = targetLeft.getAlignments().getHits(geneType);
+            List<HitMappingRecord> mRecords = new ArrayList<>(hits.length);
+            for (VDJCHit hit : hits)
+                mRecords.add(new HitMappingRecord(hit.getGene(), new Alignment[]{
+                        hit.getAlignment(targetLeft.getTargetId()), hit.getAlignment(targetRight.getTargetId())}));
+
+            // Don't resort mRecords because "hits" were already sorted. Sorting may be different from
+            // Collections.sort(mRecords, ...), if initial Alignments object contain more than two targets,
+            // however soring in "hits" is considered here to be more accurate because it was supported by
+            // other parts of the multi-read object
+            return mRecords;
+        }
+
+        // Full recalculation for targets form two different Alignments objects
+        Map<VDJCGeneId, HitMappingRecord> map = extractHitsMapping(targetLeft, targetRight, geneType);
+        List<HitMappingRecord> mRecords = new ArrayList<>(map.values());
+        Collections.sort(mRecords, new Comparator<HitMappingRecord>() {
+            @Override
+            public int compare(HitMappingRecord o1, HitMappingRecord o2) {
+                return Integer.compare(sumScore(o2.alignments), sumScore(o1.alignments));
+            }
+        });
+        return mRecords;
+    }
+
     @SuppressWarnings("unchecked")
     static Map<VDJCGeneId, HitMappingRecord> extractHitsMapping(AlignedTarget targetLeft, AlignedTarget targetRight, GeneType geneType) {
         Map<VDJCGeneId, HitMappingRecord> map = new HashMap<>();
@@ -237,18 +273,10 @@ public TargetMergingResult merge(long readId, AlignedTarget targetLeft, AlignedT
     public TargetMergingResult merge(long readId, AlignedTarget targetLeft, AlignedTarget targetRight,
                                      boolean trySequenceMerge) {
         for (GeneType geneType : GeneType.VJC_REFERENCE) {
-
-            Map<VDJCGeneId, HitMappingRecord> map = extractHitsMapping(targetLeft, targetRight, geneType);
-            if (map.isEmpty())
+            if (!hasAlignments(targetLeft, geneType) || !hasAlignments(targetRight, geneType))
                 continue;
-            List<HitMappingRecord> als = new ArrayList<>(map.values());
 
-            Collections.sort(als, new Comparator<HitMappingRecord>() {
-                @Override
-                public int compare(HitMappingRecord o1, HitMappingRecord o2) {
-                    return Integer.compare(sumScore(o2.alignments), sumScore(o1.alignments));
-                }
-            });
+            List<HitMappingRecord> als = extractSortedHits(targetLeft, targetRight, geneType);
 
             Alignment<NucleotideSequence>[] topHits = als.get(0).alignments;
 

src/main/java/com/milaboratory/mixcr/util/AlignmentExtender.java

@@ -108,6 +108,16 @@ public VDJCAlignments process(VDJCAlignments input) {
                             < vAnchorPositionInRef)
                         break OUTER;
 
+                    //checking one more time, whether extension is required
+                    //this termination point will be triggered if aligned V hits do not agree on
+                    //the position of vLeftExtensionRefPoint
+                    if (vAnchorPositionInRef >= vHit.getAlignment(cdr3target).getSequence1Range().getFrom()) {
+                        //dropping any previous extension intents
+                        vExtension = null;
+                        //breaking only current loop
+                        break;
+                    }
+
                     //extend V
                     int vLeftTargetId = -1;
                     int vLeftEndCoord = -1;
@@ -229,6 +239,16 @@ else if (!vExtension.equals(r))
                             >= jAnchorPositionInRef)
                         break OUTER;
 
+                    //checking one more time, whether extension is required
+                    //this termination point will be triggered if aligned J hits do not agree on
+                    //the position of jRightExtensionRefPoint
+                    if (jAnchorPositionInRef <= jHit.getAlignment(cdr3target).getSequence1Range().getTo()) {
+                        //dropping any previous extension intents
+                        jExtension = null;
+                        //breaking only current loop
+                        break;
+                    }
+
                     //extend J
                     int jRightTargetId = -1;
                     int jRightEndCoord = Integer.MAX_VALUE;