Merge branch 'v2'

Release 3.8.28

 - amrfinder -u option no longer overwrites databases if they already exist.
    New --force_update option will still overwrite the database as the old -u
    option did. ([#16])
 - New --protein_output and --nucleotide_output options will produce FASTA files
    containing the sequence of AMRFinderPlus hits
 - --mutation_all output has been streamlined and some bugs have been fixed
 - New --name option will prepend a "Name" field to every row of the
    AMRFinderPlus report ([#25])
 - Fusion genes are now reported with both element symbols separated by a '/' on
    both lines of the report. E.g., aac(6')-Ie/aph(2'')-If2
 - For HMM-only hits (have HMM in the method column which means they do not have
    BLAST results that meet cutoff) BLAST statistics will now be reported
    if any alignment at that location was made.
 - When in COMBINED (nucleotide + protein) mode and when an "INTERNAL_STOP" is detected
    at a locus that also has a protein, the protein result will be reported with
    Method of INTERNAL_STOP (Previously the nucleotide result was reported. This
    would affect the identity, length, and coverage statistics AMRFinderPlus reports.
 - Improved handling of partial alignments for point-mutation detection, including
    prioritization of translated matches (POINTX) when the protein match identifies fewer
    known SNPs.
 - Incorrect amino acids inferred by protein annotations extended in the 5' (N-terminal)
    direction because correct annotations start with alternative start codons will
    no longer count as mismatches.
 - Small improvements in performance.

.github/workflows/ccpp.yml

@@ -24,3 +24,12 @@ jobs:
       run: ./amrfinder -u
     - name: make test
       run: make test
+    - name: test for no-overwrite database update (PD-3469 / https://github.com/ncbi/amr/issues/16)
+      run: ./amrfinder -u 2>&1 | fgrep 'Skipping update, use amrfinder --force_update to overwrite the existing database'
+    - name: make github_binaries
+      run: make github_binaries
+    - uses: actions/upload-artifact@v2
+      with:
+          name: release-binary
+          path: amrfinder_binaries_v*.tar.gz
+

.gitignore

@@ -4,6 +4,7 @@ data/
 amr_report
 amrfinder
 fasta_check
+fasta_extract
 dna_mutation
 gff_check
 amrfinder_update

Makefile

@@ -57,7 +57,7 @@ COMPILE.cpp= $(CXX) $(CPPFLAGS) $(SVNREV) $(DBDIR) -c
 
 .PHONY: all clean install release
 
-BINARIES= amr_report amrfinder amrfinder_update fasta_check fasta2parts gff_check dna_mutation 
+BINARIES= amr_report amrfinder amrfinder_update fasta_check fasta_extract fasta2parts gff_check dna_mutation 
 
 all:	$(BINARIES)
 
@@ -89,6 +89,11 @@ fasta_checkOBJS=fasta_check.o common.o
 fasta_check:	$(fasta_checkOBJS)
 	$(CXX) -o $@ $(fasta_checkOBJS)
 
+fasta_extract.o:	common.hpp common.inc
+fasta_extractOBJS=fasta_extract.o common.o
+fasta_extract:	$(fasta_extractOBJS)
+	$(CXX) -o $@ $(fasta_extractOBJS)
+
 fasta2parts.o:	common.hpp common.inc
 fasta2partsOBJS=fasta2parts.o common.o
 fasta2parts:	$(fasta2partsOBJS)

README.md

@@ -3,7 +3,7 @@
 This software and the accompanying database are designed to find acquired antimicrobial resistance genes and some point mutations in protein or assembled nucleotide sequences. We have also added "plus" stress, head, and biocide resistance as well as some virulence factors and E. coli antigens.
 
 ## See [the wiki for documentation](https://github.com/ncbi/amr/wiki)
-[Subscribe to our announce list](https://www.ncbi.nlm.nih.gov/mailman/listinfo/amrfinder-announce) for announcements of  database or software updates.
+[Please subscribe to our announce list](https://www.ncbi.nlm.nih.gov/mailman/listinfo/amrfinder-announce) for announcements of database and software updates.
 
 ----
 # Licenses

alignment.cpp

@@ -414,9 +414,9 @@ bool normalizeSeq (const string &seq1,
   ASSERT (seq1. size () == seq2. size ());
 
   bool changed = false;
-  size_t start = NO_INDEX;
+  size_t start = no_index;
   FFOR (size_t, i, seq1. size ())
-    if (start == NO_INDEX)
+    if (start == no_index)
     {
       if (seq2 [i] == '-')
         start = i;
@@ -432,7 +432,7 @@ bool normalizeSeq (const string &seq1,
           changed = true;
         }
         else
-          start = NO_INDEX;
+          start = no_index;
       }
   ASSERT (seq1. size () == seq2. size ());
 
@@ -694,13 +694,13 @@ void Alignment::setSeqChanges (const Vector<Mutation> &refMutations,
   }
 
   // SeqChange::mutation
-	size_t start_ref_prev = NO_INDEX;
+	size_t start_ref_prev = no_index;
 	for (SeqChange& seqChange : seqChanges)
   {
     seqChange. qc ();
     if (verbose ())
       seqChange. saveText (cout);
-    IMPLY (start_ref_prev != NO_INDEX, start_ref_prev <= seqChange. start_ref);
+    IMPLY (start_ref_prev != no_index, start_ref_prev <= seqChange. start_ref);
     while (j < refMutations. size ())
     {
 		  const Mutation& mut = refMutations [j];
@@ -829,6 +829,55 @@ void Alignment::qc () const
 
 
 
+bool Alignment::getFrameShift_right (const Alignment &rightPart,
+                                     size_t diff_max) const
+{
+  ASSERT (! targetProt);
+  ASSERT (refProt);
+  ASSERT (rightPart. refProt);
+
+  if (this == & rightPart)
+    return false;
+
+  if (rightPart. targetProt)
+    return false;    
+
+  if (   targetName   != rightPart. targetName
+      || refName      != rightPart. refName
+      || targetStrand != rightPart. targetStrand
+     )
+    return false;
+
+  if (   refStart     >= rightPart. refStart
+      || refEnd       >= rightPart. refEnd
+      || refEnd + diff_max / 3 < rightPart. refStart
+     )
+    return false;
+
+  if (targetStrand)
+  {
+    if (   targetStart >= rightPart. targetStart
+        || targetEnd   >= rightPart. targetEnd
+        || targetEnd + diff_max < rightPart. targetStart
+       )
+      return false;
+  }
+  else
+  {
+    if (   targetStart <= rightPart. targetStart
+        || targetEnd   <= rightPart. targetEnd
+        || targetStart > rightPart. targetEnd + diff_max
+       )
+      return false;
+  }  
+
+  if (targetStart % 3 == rightPart. targetStart % 3)
+    return false;
+
+  return true;
+}
+
+
 
 }  // namespace
 

alignment.hpp

@@ -137,6 +137,8 @@ struct SeqChange : Root
 	char prev {'\0'};
 
 	const Mutation* mutation {nullptr};
+
+	const SeqChange* replacement {nullptr};
 
 
   SeqChange () = default;
@@ -182,6 +184,7 @@ struct SeqChange : Root
   bool finish (const string &refSeq,
                size_t flankingLen);
     // Return: good match
+    // Input: flankingLen: valid if > 0
 private:
   void setSeq ();
   void setStartStopRef ();
@@ -252,6 +255,7 @@ struct Alignment : Root
   void setSeqChanges (const Vector<Mutation> &refMutations,
                       size_t flankingLen/*,
                       bool allMutationsP*/);
+    // Input: flankingLen: valid if > 0
 public:
   bool empty () const override
     { return targetName. empty (); }
@@ -288,6 +292,19 @@ struct Alignment : Root
              && nident == refLen 
              && nident == targetSeq. size ();
 	  }
+	bool getFrameShift (const Alignment &other,
+	                    size_t diff_max) const
+	  // Return: success
+	  // Input: diff_max: in bp
+	  // Requires: !targetProt, refProt, rightPart.refProt
+	  { return    nident >= other. nident
+	           && (          getFrameShift_right (other, diff_max) 
+	               || other. getFrameShift_right (*this, diff_max) 
+	              );
+	  }
+private:
+	bool getFrameShift_right (const Alignment &rightPart,
+	                          size_t diff_max) const;
 };