Add flag to keep zero length intervals when converting bed -> interva…

…l_list (#1928)

* Add flag to keep zero length intervals when converting bed -> interval_list

src/main/java/picard/util/BedToIntervalList.java

@@ -107,9 +107,13 @@ public class BedToIntervalList extends CommandLineProgram {
     @Argument(doc = "If true, entries that are on contig-names that are missing from the provided dictionary will be dropped.")
     public boolean DROP_MISSING_CONTIGS = false;
 
+    @Argument(doc = "If true, write length zero intervals in input bed file to resulting interval list file.")
+    public boolean KEEP_LENGTH_ZERO_INTERVALS = false;
+
     private final Log LOG = Log.getInstance(getClass());
     private int missingIntervals = 0;
     private int missingRegion = 0;
+    private int lengthZeroIntervals = 0;
 
     @Override
     protected int doWork() {
@@ -168,8 +172,18 @@ protected int doWork() {
                 }
 
                 final boolean isNegativeStrand = bedFeature.getStrand() == Strand.NEGATIVE;
-                final Interval interval = new Interval(sequenceName, start, end, isNegativeStrand, name);
-                intervalList.add(interval);
+
+                // Use end+1 since bed start gets shifted by 1 using 1-based coordinates
+                if ((start == end+1) && !KEEP_LENGTH_ZERO_INTERVALS) {
+                    LOG.info(String.format("Skipping writing length zero interval at %s:%d-%d.", sequenceName, start, end));
+                } else {
+                    final Interval interval = new Interval(sequenceName, start, end, isNegativeStrand, name);
+                    intervalList.add(interval);
+                }
+
+                if (start == end+1) {
+                    lengthZeroIntervals++;
+                }
 
                 progressLogger.record(sequenceName, start);
             }
@@ -182,6 +196,19 @@ protected int doWork() {
                     LOG.warn(String.format("There were %d missing regions with a total of %d bases", missingIntervals, missingRegion));
                 }
             }
+
+            if (!KEEP_LENGTH_ZERO_INTERVALS) {
+                if (lengthZeroIntervals == 0) {
+                    LOG.info("No input regions had length zero, so none were skipped.");
+                } else {
+                    LOG.info(String.format("Skipped writing a total of %d entries with length zero in the input file.", lengthZeroIntervals));
+                }
+            } else {
+                if (lengthZeroIntervals > 0) {
+                    LOG.warn(String.format("Input file had %d entries with length zero. Run with the KEEP_LENGTH_ZERO_INTERVALS flag set to false to remove these.", lengthZeroIntervals));
+                }
+            }
+
             // Sort and write the output
             IntervalList out = intervalList;
             if (SORT) {

src/test/java/picard/util/BedToIntervalListTest.java

@@ -16,7 +16,7 @@ public class BedToIntervalListTest {
 
     private static final String TEST_DATA_DIR = "testdata/picard/util/BedToIntervalListTest";
 
-    private void doTest(final String inputBed, final String header) throws IOException, SAMException {
+    private void doTest(final String inputBed, final String header, boolean keepLengthZero) throws IOException, SAMException {
         final File outputFile  = File.createTempFile("bed_to_interval_list_test.", ".interval_list");
         outputFile.deleteOnExit();
         final BedToIntervalList program = new BedToIntervalList();
@@ -25,6 +25,7 @@ private void doTest(final String inputBed, final String header) throws IOExcepti
         program.SEQUENCE_DICTIONARY = new File(TEST_DATA_DIR, header);
         program.OUTPUT = outputFile;
         program.UNIQUE = true;
+        program.KEEP_LENGTH_ZERO_INTERVALS = keepLengthZero;
         program.doWork();
 
         // Assert they are equal
@@ -34,25 +35,30 @@ private void doTest(final String inputBed, final String header) throws IOExcepti
 
     @Test(dataProvider = "testBedToIntervalListDataProvider")
     public void testBedToIntervalList(final String inputBed) throws IOException {
-        doTest(inputBed, "header.sam");
+        doTest(inputBed, "header.sam", true);
     }
 
     // test a fixed bed file using different dictionaries
     @Test(dataProvider = "testBedToIntervalListSequenceDictionaryDataProvider")
     public void testBedToIntervalListSequenceDictionary(final String dictionary) throws IOException {
-        doTest("seq_dict_test.bed", dictionary);
+        doTest("seq_dict_test.bed", dictionary, true);
     }
 
     // test for back dictionaries - we expect these to throw exceptions
     @Test(dataProvider = "testBedToIntervalListSequenceDictionaryBadDataProvider",
           expectedExceptions = {SAMException.class, PicardException.class})
     public void testBedToIntervalListBadSequenceDictionary(final String dictionary) throws IOException {
-        doTest("seq_dict_test.bed", dictionary);
+        doTest("seq_dict_test.bed", dictionary, true);
     }
 
     @Test(dataProvider = "testBedToIntervalListOutOfBoundsDataProvider", expectedExceptions = PicardException.class)
     public void testBedToIntervalListOutOfBounds(final String inputBed) throws IOException {
-        doTest(inputBed, "header.sam");
+        doTest(inputBed, "header.sam", true);
+    }
+
+    @Test(dataProvider = "testLengthZeroIntervalsSkippedProvider")
+    public void testLengthZeroIntervalsSkipped(final String inputBed) throws IOException {
+        doTest(inputBed, "header.sam", false);
     }
 
     @DataProvider
@@ -103,4 +109,11 @@ public Object[][] testBedToIntervalListOutOfBoundsDataProvider() {
                 {"off_by_one_interval.bed"}
         };
     }
+
+    @DataProvider
+    public Object[][] testLengthZeroIntervalsSkippedProvider() {
+        return new Object[][]{
+                {"zero_length_test.bed"}
+        };
+    }
 }

testdata/picard/util/BedToIntervalListTest/zero_length_test.bed

@@ -0,0 +1,4 @@
+chr1	1000	1000
+chr2	1000	2000
+chr3	1000	1000
+chr4	1000	2000

testdata/picard/util/BedToIntervalListTest/zero_length_test.bed.interval_list

@@ -0,0 +1,11 @@
+@HD	VN:1.6	SO:coordinate
+@SQ	SN:chr1	LN:1000000
+@SQ	SN:chr2	LN:1000000
+@SQ	SN:chr3	LN:1000000
+@SQ	SN:chr4	LN:1000000
+@SQ	SN:chr5	LN:1000000
+@SQ	SN:chr6	LN:1000000
+@SQ	SN:chr7	LN:1000000
+@SQ	SN:chr8	LN:1000000
+chr2	1001	2000	+	.
+chr4	1001	2000	+	.