Skip to content

Commit

Permalink
Redux: dup group clean-up only
Browse files Browse the repository at this point in the history
  • Loading branch information
@charlesshale
charlesshale committed Sep 9, 2024
1 parent 0aa3f00 commit f7c550a
Showing 1 changed file with 8 additions and 109 deletions.
117 changes: 8 additions & 109 deletions redux/src/main/java/com/hartwig/hmftools/redux/common/DuplicateGroup.java
View file
Original file line number Diff line number Diff line change
Expand Up @@ -49,15 +49,11 @@ public class DuplicateGroup
private final Map<String,Boolean> mReadIdInitialIsFirst;
private ReadTypeId mInitialPrimaryDetails;

@Deprecated
private final ReadTypeId[] mPrimaryReadTypeIndex; // details for primary and mate reads

private TemplateReadData mPrimaryTemplateRead; // read data from the primary consensus read
private String mGroupReadId;
private boolean mDualStrand;

private static final int MAX_READ_TYPES = ReadType.values().length;
private static final int PRIMARY_READ_TYPES = ReadType.MATE.ordinal() + 1;

public DuplicateGroup(final String id, final Fragment fragment)
{
Expand All @@ -76,8 +72,6 @@ public DuplicateGroup(final String id, final Fragment fragment)
mPrimaryTemplateRead = null;
mGroupReadId = null;
mDualStrand = false;

mPrimaryReadTypeIndex = new ReadTypeId[PRIMARY_READ_TYPES];
}

public List<Fragment> fragments() { return mFragments; }
Expand Down Expand Up @@ -143,10 +137,8 @@ else if(!read.getSupplementaryAlignmentFlag() && hasValidSupp)
fragment.setUmi(mUmiId != null ? mUmiId : "");
fragment.setStatus(DUPLICATE);

boolean checkFirstInPair = fragment.isPreciseInversion();

// add non-supps first to establish the correct primary read type info
fragment.reads().stream().filter(x -> !x.getSupplementaryAlignmentFlag()).forEach(x -> addRead(x, checkFirstInPair));
fragment.reads().stream().filter(x -> !x.getSupplementaryAlignmentFlag()).forEach(x -> addRead(x));
}

// add supplementaries once all primaries have been added and their expected supps & types registered
Expand All @@ -155,12 +147,9 @@ else if(!read.getSupplementaryAlignmentFlag() && hasValidSupp)
mFragments.clear();
}

public void addRead(final SAMRecord read) { addRead(read, false); }

private void addRead(final SAMRecord read, boolean checkFirstInPair)
public void addRead(final SAMRecord read)
{
int readTypeIndex = getReadTypeIndex(read);
// int readTypeIndex = getReadTypeIndexOld(read, checkFirstInPair);

if(mReadGroups[readTypeIndex] == null)
mReadGroups[readTypeIndex] = Lists.newArrayListWithExpectedSize(mFragmentCount);
Expand Down Expand Up @@ -199,31 +188,21 @@ private class ReadTypeId
// so coordinates are also used
public final String Chromosome;
public final int UnclippedPosition;
public final int[] PositionRange;
public final byte Orientation;
public final boolean FirstInPair;
public boolean HasSupplementary;
public final boolean Unmapped;

public ReadTypeId(
final String chromosome, final int unclippedPosition, final int position, final byte orientation,
final boolean hasSupplementary, final boolean firstInPair, final boolean unmapped)
final String chromosome, final int unclippedPosition, final byte orientation,
final boolean firstInPair, final boolean unmapped)
{
Chromosome = chromosome;
UnclippedPosition = unclippedPosition;
PositionRange = new int[] { position, position };
Orientation = orientation;
HasSupplementary = hasSupplementary;
FirstInPair = firstInPair;
Unmapped = unmapped;
}

public void updatePosition(final int position)
{
PositionRange[SE_START] = min(PositionRange[SE_START], position);
PositionRange[SE_END] = max(PositionRange[SE_END], position);
}

private boolean primaryMatches(final SAMRecord read, boolean checkFirstInPair)
{
if(Unmapped || read.getReadUnmappedFlag())
Expand All @@ -242,23 +221,9 @@ private boolean primaryMatches(final SAMRecord read, boolean checkFirstInPair)
return FirstInPair == read.getFirstOfPairFlag();
}

public boolean supplementaryMatches(final SAMRecord read, final SupplementaryReadData suppData)
{
if(!suppData.Chromosome.equals(Chromosome))
return false;

if(orientation(read) != Orientation)
return false;

return positionWithin(suppData.Position, PositionRange[SE_START], PositionRange[SE_END]);
}

public String toString()
{
return format("%s:%d%s%s %d-%d %s",
Chromosome, UnclippedPosition, Orientation == POS_ORIENT ? " " : "_R ",
FirstInPair ? "R1" : "R2", PositionRange[SE_START], PositionRange[SE_END],
HasSupplementary ? "has-supp" : "");
return format("%s:%d%s%s", Chromosome, UnclippedPosition, Orientation == POS_ORIENT ? " " : "_R ", FirstInPair ? "R1" : "R2");
}
}

Expand Down Expand Up @@ -295,11 +260,10 @@ private boolean matchesInitialPrimaryRead(final SAMRecord read)
// must be a primary, not supplementary read
if(mInitialPrimaryDetails == null)
{
boolean isUnmapped = read.getReadUnmappedFlag();
mInitialPrimaryDetails = new ReadTypeId(
read.getReferenceName(),
read.getReadUnmappedFlag() ? 0 : getFivePrimeUnclippedPosition(read), read.getAlignmentStart(),
read.getReadUnmappedFlag() ? 0 : orientation(read),
false, read.getFirstOfPairFlag(), read.getReadUnmappedFlag());
read.getReferenceName(), isUnmapped ? 0 : getFivePrimeUnclippedPosition(read),
isUnmapped ? 0 : orientation(read), read.getFirstOfPairFlag(), isUnmapped);

initialIsFirst = read.getFirstOfPairFlag();
}
Expand All @@ -316,71 +280,6 @@ private boolean matchesInitialPrimaryRead(final SAMRecord read)
return initialIsFirst == read.getFirstOfPairFlag();
}

private int getReadTypeIndexOld(final SAMRecord read, boolean checkFirstInPair)
{
if(!read.getSupplementaryAlignmentFlag())
{
SupplementaryReadData suppData = SupplementaryReadData.extractAlignment(read);
boolean hasValidSupp = suppData != null && HumanChromosome.contains(suppData.Chromosome);

int index = 0;
for(; index < mPrimaryReadTypeIndex.length; ++index)
{
if(mPrimaryReadTypeIndex[index] == null)
break;

if(mPrimaryReadTypeIndex[index].primaryMatches(read, checkFirstInPair))
{
// update details for this primary index
mPrimaryReadTypeIndex[index].updatePosition(read.getAlignmentStart());
mPrimaryReadTypeIndex[index].HasSupplementary |= hasValidSupp;
return index;
}
}

if(index >= mPrimaryReadTypeIndex.length)
{
RD_LOGGER.error("group({}) non-supp read unmatched: {}", this, readToString(read));
return index;
}

mPrimaryReadTypeIndex[index] = new ReadTypeId(
read.getReferenceName(),
read.getReadUnmappedFlag() ? 0 : getFivePrimeUnclippedPosition(read), read.getAlignmentStart(),
read.getReadUnmappedFlag() ? 0 : orientation(read),
hasValidSupp, read.getFirstOfPairFlag(), read.getReadUnmappedFlag());

// register the expected supplementary read group
if(hasValidSupp)
{
int suppReadTypeIndex = index == ReadType.INITIAL_PRIMARY.ordinal() ?
ReadType.INITIAL_SUPPLEMENTARY.ordinal() : ReadType.MATE_SUPPLEMENTARY.ordinal();

if(mReadGroups[suppReadTypeIndex] == null)
mReadGroups[suppReadTypeIndex] = Lists.newArrayListWithExpectedSize(mFragmentCount);
}

return index;
}

SupplementaryReadData suppData = SupplementaryReadData.extractAlignment(read);

int matchedPrimaryIndex;

if(suppData == null)
{
// logical assert since all supp should have this attribute set
matchedPrimaryIndex = mPrimaryReadTypeIndex[0] != null && mPrimaryReadTypeIndex[0].HasSupplementary ? 0 : 1;
}
else
{
// match the supplementary details to that of the primary read
matchedPrimaryIndex = mPrimaryReadTypeIndex[0] != null && mPrimaryReadTypeIndex[0].supplementaryMatches(read, suppData) ? 0 : 1;
}

return matchedPrimaryIndex == 0 ? ReadType.INITIAL_SUPPLEMENTARY.ordinal() : ReadType.MATE_SUPPLEMENTARY.ordinal();
}

public boolean allReadsReceived()
{
for(int i = 0; i < mReadGroups.length; ++i)
Expand Down

0 comments on commit f7c550a

Please sign in to comment.