public void TestIfBamIsStitched()
{
//test some generic bam
var extractor = new BamFileAlignmentExtractor(Path.Combine(TestPaths.LocalTestDataDirectory, "unaligned.bam"));
Assert.Equal(false, extractor.SourceIsStitched);
//test to be robust to crazy bams.
Assert.Equal(false,
BamFileAlignmentExtractor.CheckBamHeaderIfBamHasBeenStitched(""));
Assert.Equal(false,
BamFileAlignmentExtractor.CheckBamHeaderIfBamHasBeenStitched("@PG @PG"));
Assert.Equal(false,
BamFileAlignmentExtractor.CheckBamHeaderIfBamHasBeenStitched("blah"));
Assert.Equal(false,
BamFileAlignmentExtractor.CheckBamHeaderIfBamHasBeenStitched(null));
//test some real normal headers
Assert.Equal(true,
BamFileAlignmentExtractor.CheckBamHeaderIfBamHasBeenStitched(GetPiscesStitchedHeader()));
Assert.Equal(false,
BamFileAlignmentExtractor.CheckBamHeaderIfBamHasBeenStitched(GetRegularHeader()));
}
public void IntervalJumping_Ends()
{
var smallBam = Path.Combine(UnitTestPaths.TestDataDirectory, "Ins-L3-var12_S12.bam");
var intervals = new Dictionary <string, List <Region> >();
var chrIntervals = new List <Region>
{
new Region(28608100, 28608100), // interval in the middle of coverage
new Region(29608700, 29608800) // interval out in the boonies where there's no data
};
intervals.Add("chr13", chrIntervals);
var extractor = new BamFileAlignmentExtractor(smallBam, bamIntervals: intervals);
var read = new Read();
var numReadsLessThan = 0;
var numReadsGreaterThan = 0;
// verify we are always moving forward and not backwards (not re-reading alignments)
while (extractor.GetNextAlignment(read))
{
if (read.EndPosition + 1 < 28608100) // bam reader is off by one, see note in BamFileExtractor.Jump
{
numReadsLessThan++;
}
else if (read.Position > 28608100)
{
numReadsGreaterThan++;
}
}
Assert.Equal(1, numReadsLessThan); // this should be just the first read (before we figure out we're not in range)
Assert.Equal(0, numReadsGreaterThan);
}
public void SanityCheckSequenceOrdering()
{
var smallBam = Path.Combine(TestPaths.LocalTestDataDirectory, "Ins-L3-var12_S12.bam");
var intervals = new Dictionary <string, List <Region> >();
var chrIntervals = new List <Region>
{
new Region(28607838, 28607838),
new Region(28608631, 28608631)
};
var expectedSQorder = new List <string> {
"chr10", "chr11", "chr12", "chr13"
}; //I dont know why it starts with 10, thats just how it is in the bam. thats what makes it a good test case.
intervals.Add("chr13", chrIntervals);
var extractor = new BamFileAlignmentExtractor(smallBam);
List <string> sequencesInTheBamOrder = extractor.SourceReferenceList;
Assert.Equal(expectedSQorder[0], sequencesInTheBamOrder[0]);
Assert.Equal(expectedSQorder[1], sequencesInTheBamOrder[1]);
Assert.Equal(expectedSQorder[3], sequencesInTheBamOrder[3]);
Assert.Equal(25, sequencesInTheBamOrder.Count);
//happyPath
Assert.False(extractor.SequenceOrderingIsNotConsistent(new List <string> {
"chr1", "chr2"
}));
Assert.False(extractor.SequenceOrderingIsNotConsistent(new List <string> {
"chr1", "chr3", "chr4"
}));
Assert.False(extractor.SequenceOrderingIsNotConsistent(new List <string> {
"chr14", "chr9"
})); //only b/c the bam header is silly.
//not OK
Assert.True(extractor.SequenceOrderingIsNotConsistent(new List <string> {
"chr2", "chr1"
}));
Assert.True(extractor.SequenceOrderingIsNotConsistent(new List <string> {
"chr9", "chr14"
}));
Assert.True(extractor.SequenceOrderingIsNotConsistent(new List <string> {
"chr22", "chr21"
}));
//genome has chr not in bam, be ok with it
Assert.False(extractor.SequenceOrderingIsNotConsistent(new List <string> {
"chr1", "chrMotherGoose"
}));
//bam has chr not in genome, be ok with it
Assert.False(extractor.SequenceOrderingIsNotConsistent(new List <string> {
"chr1"
}));
//empty lists
Assert.False(extractor.SequenceOrderingIsNotConsistent(new List <string> {
}));
Assert.False(extractor.SequenceOrderingIsNotConsistent(null));
}
protected virtual IAlignmentSource CreateAlignmentSource(ChrReference chrReference, string bamFilePath)
{
var alignmentExtractor = new BamFileAlignmentExtractor(bamFilePath, _options.StitchReads, chrReference.Name);
var mateFinder = _options.StitchReads ? new AlignmentMateFinder(Constants.MaxFragmentSize) : null; // jg todo - do we want to expose this to command line?
var stitcher = _options.StitchReads ? CreateStitcher() : null;
var config = new AlignmentSourceConfig
{
MinimumMapQuality = _options.MinimumMapQuality,
OnlyUseProperPairs = _options.OnlyUseProperPairs,
};
return(new AlignmentSource(alignmentExtractor, mateFinder, stitcher, config));
}
public void UnalignedReads()
{
var extractor = new BamFileAlignmentExtractor(Path.Combine(TestPaths.LocalTestDataDirectory, "unaligned.bam"));
var read = new Read();
var count = 0;
while (extractor.GetNextAlignment(read))
{
count++;
}
Assert.Equal(138826, count);
Assert.Equal(null, read.Chromosome); // last reads are unaligned
}
public void ReadFileAsStitched()
{
var smallBam = Path.Combine(TestPaths.LocalTestDataDirectory, "small.bam");
//we claim its not stitched
var extractor = new BamFileAlignmentExtractor(smallBam, false);
var read = new Read();
extractor.GetNextAlignment(read);
Assert.False(extractor.SourceIsStitched);
//we claim its stitched
extractor = new BamFileAlignmentExtractor(smallBam, true);
read = new Read();
extractor.GetNextAlignment(read);
Assert.True(extractor.SourceIsStitched);
}
public void IntervalJumping_Boundaries()
{
var smallBam = Path.Combine(UnitTestPaths.TestDataDirectory, "Ins-L3-var12_S12.bam");
var intervals = new Dictionary <string, List <Region> >();
var chrIntervals = new List <Region>
{
new Region(115169880, 115169880) // feeding in an interval that's past reference max shouldnt cause it to blow up
};
intervals.Add("chr13", chrIntervals);
var extractor = new BamFileAlignmentExtractor(smallBam, bamIntervals: intervals);
var read = new Read();
while (extractor.GetNextAlignment(read))
{
}
}
public void IntervalJumping_Middle()
{
var smallBam = Path.Combine(UnitTestPaths.TestDataDirectory, "Ins-L3-var12_S12.bam");
var intervals = new Dictionary <string, List <Region> >();
var chrIntervals = new List <Region>
{
new Region(28607838, 28607838),
new Region(28608631, 28608631)
};
intervals.Add("chr13", chrIntervals);
var extractor = new BamFileAlignmentExtractor(smallBam, bamIntervals: intervals);
var read = new Read();
// verify we skip over the middle
while (extractor.GetNextAlignment(read))
{
Assert.True(read.Position < 28607840 || read.BamAlignment.GetEndPosition() >= 28608631);
}
}
protected virtual IAlignmentSource CreateAlignmentSource(ChrReference chrReference, string bamFilePath, List <string> chrsToProcess = null)
{
AlignmentMateFinder mateFinder = null;
var alignmentExtractor = new BamFileAlignmentExtractor(bamFilePath, chrReference.Name, _bamIntervalLookup.ContainsKey(bamFilePath) && _options.SkipNonIntervalAlignments ? _bamIntervalLookup[bamFilePath] : null);
//Warn if the bam has sequences ordered differently to the reference genome.
//That would confuse us because we will not know how the user wants to order the output gvcf.
if (alignmentExtractor.SequenceOrderingIsNotConsistent(chrsToProcess))
{
Logger.WriteToLog("Warning: Reference sequences in the bam do not match the order of the reference sequences in the genome. Check bam " + bamFilePath);
Logger.WriteToLog("Variants will be ordered according to the reference genome");
}
var config = new AlignmentSourceConfig
{
MinimumMapQuality = _options.MinimumMapQuality,
OnlyUseProperPairs = _options.OnlyUseProperPairs,
};
return(new AlignmentSource(alignmentExtractor, mateFinder, config));
}
protected virtual IAlignmentSource CreateAlignmentSource(ChrReference chrReference, string bamFilePath, bool commandLineSaysStitched, List <string> chrsToProcess = null)
{
AlignmentMateFinder mateFinder = null;
var alignmentExtractor = new BamFileAlignmentExtractor(bamFilePath, commandLineSaysStitched, chrReference.Name);
//Warn if the bam has sequences ordered differently to the reference genome.
//That would confuse us because we will not know how the user wants to order the output gvcf.
if (alignmentExtractor.SequenceOrderingIsNotConsistent(chrsToProcess))
{
Logger.WriteToLog("Warning: Reference sequences in the bam do not match the order of the reference sequences in the genome. Check bam " + bamFilePath);
Logger.WriteToLog("Variants will be ordered according to the reference genome");
}
var config = new AlignmentSourceConfig
{
MinimumMapQuality = _options.BamFilterParameters.MinimumMapQuality,
OnlyUseProperPairs = _options.BamFilterParameters.OnlyUseProperPairs,
SkipDuplicates = _options.BamFilterParameters.RemoveDuplicates
};
return(new AlignmentSource(alignmentExtractor, mateFinder, config));
}
private void ReadFileTest(string bamfile, int expectedReads, bool bamHasXc)
{
var extractor = new BamFileAlignmentExtractor(bamfile);
var read = new Read();
var lastPosition = -1;
var numReads = 0;
bool hasAnyStitchedCigars = false;
while (extractor.GetNextAlignment(read))
{
Assert.True(read.Position >= lastPosition); // make sure reads are read in order
Assert.False(string.IsNullOrEmpty(read.Name));
Assert.False(string.IsNullOrEmpty(read.Chromosome));
if (!bamHasXc)
{
Assert.Equal(null, read.StitchedCigar);
}
if (read.StitchedCigar != null && read.StitchedCigar.Count > 0)
{
hasAnyStitchedCigars = true;
}
lastPosition = read.Position;
numReads++;
}
if (bamHasXc)
{
Assert.True(hasAnyStitchedCigars);
}
Assert.Equal(expectedReads, numReads);
extractor.Dispose();
// make sure can't read after dispose
Assert.Throws <IOException>(() => extractor.GetNextAlignment(read));
}
public void IntervalJumping_SmallIntervals()
{
var smallBam = Path.Combine(UnitTestPaths.TestDataDirectory, "Ins-L3-var12_S12.bam");
var intervals = new Dictionary <string, List <Region> >();
var chrIntervals = new List <Region>
{
new Region(28607838, 28607838),
new Region(28607908, 28607908),
new Region(28608631, 28608631)
};
intervals.Add("chr13", chrIntervals);
var extractor = new BamFileAlignmentExtractor(smallBam, bamIntervals: intervals);
var read = new Read();
var lastPosition = 0;
// verify we are always moving forward and not backwards (not re-reading alignments)
while (extractor.GetNextAlignment(read))
{
Assert.True(read.Position >= lastPosition);
lastPosition = read.Position;
}
}