1. Uses blocked gzip compression.
2. No more -bzip option available (since we can't compress to sdout).
3. Only file extensions that are compressed are .gz and .gzip.
4. No more need for CompressedVCFWriter.java
git-svn-id: file:///humgen/gsa-scr1/gsa-engineering/svn_contents/trunk@4099 348d0f76-0448-11de-a6fe-93d51630548a
We can now emit bzipped VCFs from the GATK.
Details: any walker that defines a VCFWriter for its @Output (i.e. pretty much every core walker from UG and on), also has associated with it the -bzip (--bzip_compression) boolean argument. When set, it will emit a VCF that is compressed with bzip2.
git-svn-id: file:///humgen/gsa-scr1/gsa-engineering/svn_contents/trunk@4093 348d0f76-0448-11de-a6fe-93d51630548a
is an output stream, we'll automatically create it and point it to stdout. Otherwise,
we'll leave it empty.
I think about it like this: marking a field 'required' indicates to the GATK that the
walker author requires a value for this field, and if the GATK can provide one without
end user intervention, it will. Maybe this is hackish. We'll try it and see.
git-svn-id: file:///humgen/gsa-scr1/gsa-engineering/svn_contents/trunk@4091 348d0f76-0448-11de-a6fe-93d51630548a
most appropriate based on feature density, and the longest feature in the file. Also:
- Converted Tribble to TestNG; it has better features and is about 6x faster.
- As much code clean-up as I could get done. More to do, especially in the example code.
- Moved asserts in the code to throw exceptions.
- Added getBinSize to the index interface; both indexes already implemented this.
- Removed the abstract parts of the indexCreator interface; this is now more simple.
- Added an IndexType enumeration; might be overkill but it is at least a single point of entry for index information.
git-svn-id: file:///humgen/gsa-scr1/gsa-engineering/svn_contents/trunk@4082 348d0f76-0448-11de-a6fe-93d51630548a
-o/--out instead of -varout. Let's watch as our intrepid support engineer
gracefully responds to all the incoming questions of the form: "the GATK told
me to use -o instead of -varout. What do I do?"
git-svn-id: file:///humgen/gsa-scr1/gsa-engineering/svn_contents/trunk@4078 348d0f76-0448-11de-a6fe-93d51630548a
Thanks to Matt for the advice.
Also, for Guillermo: while I was at it, I changed the .stats debug output to emit the original interval instead of the cleaned region.
git-svn-id: file:///humgen/gsa-scr1/gsa-engineering/svn_contents/trunk@4058 348d0f76-0448-11de-a6fe-93d51630548a
While I was fixing it, I noticed that we weren't allowing the cleaner to un-clean reads with indels when they're wrong even though we should. Hypothetically, that should rarely happen: only when we can left-align out an indel or when the original mapper really went haywire. This situation is rare enough that I'm calling logger.info to let the user know it's happening and suggesting that they double-check that everything looks right with their reads. Better to be extra-cautious now that the cleaner is moving into the 1kg and Broad production pipelines soon
.
Mark, have no fear: this was truly a rare edge case - one that won't affect the cleaning stats. There is no need to re-clean the data processing paper bams!
git-svn-id: file:///humgen/gsa-scr1/gsa-engineering/svn_contents/trunk@4057 348d0f76-0448-11de-a6fe-93d51630548a
Some refactoring of Queue extensions for reusability in scripts.
Putting the extensions into the Queue.jar after building them.
More updates to GATK walker arguments specifying @Input and @Output for Queue.
git-svn-id: file:///humgen/gsa-scr1/gsa-engineering/svn_contents/trunk@4032 348d0f76-0448-11de-a6fe-93d51630548a
- GATKVCFWriter deleted, to be replaced if absolutely necessary when VCF writing goes into Tribble.
- VCFWriter is now an interface, for easier redirection.
- VCFWriterImpl fleshes out the VCFWriter interface.
git-svn-id: file:///humgen/gsa-scr1/gsa-engineering/svn_contents/trunk@4026 348d0f76-0448-11de-a6fe-93d51630548a
Added ability to skip up-to-date jobs where the outputs are older than the inputs.
Changed -T CountDuplicates --quiet to --quietLocus so that Queue GATK extensions can use both short and full argument names.
Short names can be used to set values on Queue GATK extensions, for example: vf.XL :+= myFile
Moved Hidden from the GATK to StingUtils.
Updated ivy from 2.0.0 to 2.2.0-rc1 to fix sha1 issue: http://bit.ly/aX72w7
Added Queue to javadoc and testing build targets.
Added first Queue unit test.
Another pass at avoiding cycles in the DAG thanks to all function I/O being files.
git-svn-id: file:///humgen/gsa-scr1/gsa-engineering/svn_contents/trunk@4017 348d0f76-0448-11de-a6fe-93d51630548a
- Eliminate reduncancy of filter application.
- Track filter metrics per-shard to facitate per merging.
- Flatten counting iterator hierarchy for easier debugging.
- Rename Reads class to ReadProperties and track it outside of the Sting iterators.
Note: because shards are currently tied so closely to reads and not the merged triplet of <reads,ref,RODs>, the metrics
classes are managed by the SAMDataSource when they should be managed by something more general. For now, we're hacking
the reads data source to manage the metrics; in the future, something more general should manage the metrics classes.
git-svn-id: file:///humgen/gsa-scr1/gsa-engineering/svn_contents/trunk@4015 348d0f76-0448-11de-a6fe-93d51630548a
rpoplin
delangel
ebanks
hanna
aaron
kiran
corin
depristo
kshakir
flannick