b33873206aGATKRunReport now has an ID (random 32 char string) that uniquely identifies the JOB run and can be used to find a run in the run repository
depristo
2010-08-31 16:18:57 +0000
5e710050d6minor change, bamFiles comes from the input list, not the script
chartl
2010-08-31 16:03:35 +0000
1a14dbee1eAdding in .bam indexing; commit for Khalid
chartl
2010-08-31 15:21:41 +0000
3c956110f3Fixing up the VCFWriter storage code: instead of assuming all samples are coming from the input bam file (they're not), just use the original VCF header for writing the temporary thread files. Now parallelization in e.g. the Genomic Annotator works.
ebanks
2010-08-31 02:16:07 +0000
69d92fab4fadding the ability to get iterators from Tribble without having an index, and updating the Tabix code to the latest Samtools SVN version (this still doesn't fix the outstanding tabix bugs, waiting for Heng on that).
aaron
2010-08-30 21:49:23 +0000
e64d1be475Check if VC is null before trying to subset it (can happen with indels)
chartl
2010-08-30 20:43:37 +0000
e14a347e2eNow prints cluster report to a single PDF, rather than a dozen different PDFs.
kiran
2010-08-29 18:58:39 +0000
1ddb5d17c9hostname now fully qualified and working
depristo
2010-08-29 17:04:37 +0000
9556004dbbnow supports -o option as well as verbose output mode
depristo
2010-08-29 16:00:00 +0000
4c28fc3a39Clear documentation for GATKRunReport
depristo
2010-08-29 15:59:25 +0000
16b75e3b9aA new version of the ErrorRateByReadPosition walker, using the GATKReport functionality to store and emit its output. This version of the walker is roughly half the number of lines as the previous version, owing simply to the removal of all of the output formatting that's now handled by GATKReport.
kiran
2010-08-29 05:41:13 +0000
fd19c63aafA data structure that allows data to be collected over the course of a walker's computation, then have that data written to a PrintStream such that it's human-readable, AWK-able, and R-friendly (given that you load it using the GATKReport loader module).
kiran
2010-08-29 05:39:24 +0000
df76474b34Proper filtering when indels are being lifted over
ebanks
2010-08-29 04:48:31 +0000
2ffa98aea5Ugh! varout --> out
chartl
2010-08-29 02:34:41 +0000
d7edce31a2Commit of fCP for Khalid
chartl
2010-08-29 02:24:25 +0000
3fd2392090Improved interface to getting command line options. Now fully traverses all objects to get all internal argument collections. Preliminary (but disabled version) of phoning home (see -et argument for more information). Captures correct and erroring out runs and writes out gzipped, xml report with lots of useful information. Needs a bit more information but is approximately working. Reports going to /humgen/gsa-hpprojects/GATK/reports/ in submitted directory that will be collated by some external tool. Only operating if -et STANDARD or -et STDOUT are provided currently and REPORT_DIR contains a file called ENABLE. WalkerTest now adds -et NO_ET to tests to avoid populating the reports with tests.
depristo
2010-08-28 22:53:32 +0000
576ae30df1A version of the full calling pipeline queue script that fully compiles without String/File/NamedFile type exceptions (e.g. expected String but got NamedFile/Expected NamedFile but got File). Pipeline itself is under testing with 5 bam files.
chartl
2010-08-28 22:51:11 +0000
9c3f403307Add the calculated lod value to the info field of each recalibrated VCF record.
rpoplin
2010-08-27 21:33:58 +0000
fe19539188Small bug fix: if a read falls at the edge of an indel event (but is not part of it), don't count it towards consistency computation.
delangel
2010-08-27 20:37:27 +0000
54355b1864In variant quality score recalibrator Preserve the definition of known and novel to be presence in dbSNP or not even when training with 1KG project calls.
rpoplin
2010-08-27 19:07:59 +0000
7a5f297083actually modify the vcf when a sample has been down-sampled
ebanks
2010-08-27 19:03:21 +0000
c6441b585aActually hook up the new indel genotyper and merge analyses into DAG (aka "i forgot to add()")
chartl
2010-08-27 18:00:50 +0000
9860db64a3Fix up liftover to enable lifting over indels
ebanks
2010-08-27 17:55:27 +0000
fb177c4feeIf only dcov is specified, assume that selected downsample type is BY_SAMPLE.
hanna
2010-08-27 17:35:41 +0000
9584cbc05eUG now downsamples to 250x by default
ebanks
2010-08-27 16:53:15 +0000
431392330eRe-enable the max records in ram argument, which I accidentally removed
ebanks
2010-08-27 16:42:49 +0000
7908237b90Full calling pipeline now calls indels through the indel genotyper, merges with combine variants, and filters on them. Since new genomic annotator is fast, it is no longer scatter-gathered.
chartl
2010-08-27 16:28:24 +0000
78946c4ffdAllowing the Queue to run the GATK via -cp instead of only from -jar. Added an example of using a walker with Queue and a custom -classpath. Removed an unused import statement in NamedFileWrapper.
kshakir
2010-08-27 16:25:59 +0000
de5ccfb0b1Moved hasPileupBeenDownsampled() based on Eric's request. Also eliminated @Deprecated constructors from AlignmentContext.
hanna
2010-08-27 16:12:05 +0000
427a2f85e9The Indel Realigner now lets the engine do all of the setup for args affecting the SAM writer. Thanks, Matt!
ebanks
2010-08-27 15:19:47 +0000
a3d9d23b0fNow prints het genotype with GQ=0 for each indel; in two-sample (normal-tumor) mode, prints both genotypes (N and T) as hets for germline events or hom ref for N and het for T for somatic events (all genotypes still have GQ=0)
asivache
2010-08-27 15:06:42 +0000
dda84a0e54Re-enabling indels for the Genomic Annotator as per Steve's patch. Steve assures me that he will test this out really well.
ebanks
2010-08-27 15:01:25 +0000
6f4af47aacsetMaxRecordsInRam now a member of StingSAMFileWriter.
hanna
2010-08-27 14:50:41 +0000
467405094aup the test mem. from 2g to 4g; we're currently hitting the 2g in aggregate across some of the larger tests
aaron
2010-08-27 01:39:05 +0000
bfcac33e80Cleaning up playground utils and tests
ebanks
2010-08-27 01:25:47 +0000
4979dcc9a7Finishing up the playground cleanup (for now)
ebanks
2010-08-27 01:19:37 +0000
0452b1ab68archiving, removing, or promoting to core from playground
ebanks
2010-08-27 01:07:42 +0000
d773b3264bEliminated -mrl option. Eliminated -fmq0 option. Eliminated read group hallucination.
hanna
2010-08-26 21:38:03 +0000
7671502e1bChanges from James Pirruccello: now can handle differences between UCSC and NCBI tables, properly sorting despite the contig prefix differences (presence or absence of 'chr'), and converts NCBI format to UCSC format for use by the GenomicAnnotator.
kiran
2010-08-26 19:02:29 +0000
8931a63588updated a whole bunch of column names to work like i want them to and added more informative figures for DOC
corin
2010-08-26 18:19:09 +0000
f384d4a5d6A java reimplementation of vcf2table in python; supports getting more useful information about genotypes (HET, e.g.) than was possible in python.
depristo
2010-08-26 17:50:33 +0000
1e193e4c20prinring '\n' at the end of line leads to some aesthetical advantages
asivache
2010-08-26 16:29:42 +0000
9b3ffa5f64Now outputs VCF (as standard output associated with -o)! Can also outptut, in parallel, a lightweight bed and fully annotated .txt (old verbose format) with --bed and --verbose, respectively
asivache
2010-08-26 16:26:03 +0000
dfae48cee0Moving supported tools to core
ebanks
2010-08-26 13:56:19 +0000
45d895dcf4Remove the check in the Unified Genotyper for hitting the max reads at locus value. Instead, simply add a flag to the INFO field if any of the samples has been downsampled. 95% hooked up.
ebanks
2010-08-26 05:50:47 +0000
e06b2c90efCap the default size of join tables; this can be modified with the --maxJoinTableSize argument. Also, misc cleanup of the comments.
ebanks
2010-08-26 05:21:26 +0000
79cd716671More cleanup of the Genomic Annotator. Also, we now require join tables to have unique entries for the column keyed on the join.
ebanks
2010-08-26 04:43:52 +0000
0105e8d063Updated Queue GATK generation to reflect -B and -I changes.
kshakir
2010-08-25 22:17:36 +0000
bdb3a7ebe6The tagger was automatically combining identical tags, but this is a problem for the ROD system. Eliminate tag combine operation.
hanna
2010-08-25 22:01:32 +0000
39da567d48Changed ReadBackedPhasing to be a RodWalker (corrected to By(READS))
fromer
2010-08-25 20:53:04 +0000
a7af605d95update to use new rod syntax
ebanks
2010-08-25 20:21:53 +0000
4678613893Significant fixes for the Genomic Annotator. 1. Rip out all of Ben's code intended to circumvent the stable VCF Writer output system in multi-threaded mode (I threw up a little when I saw this code). This will improve memory consumption when running with -nt. 2. Don't annotate indels or > bi-allelic sites. 3. Fix bug where not all records were making it into the output VCF. 4. General code clean up.
ebanks
2010-08-25 20:16:50 +0000
41e53d37e1Changed ReadBackedPhasing to be a RodWalker (more efficient, since it is ROD-focused)
fromer
2010-08-25 19:43:57 +0000
6eb1559c1dEnd-to-end calling works again (changes to walker arguments, and changes to queue, affect its validity, so it often goes out-of-date before I try to use it again)
chartl
2010-08-25 18:52:44 +0000
fba71e3c15Placeholder commit. Implements a loader for a new multi-part GATK reporting format. See what it looks like at /home/radon01/kiran/scr1/projects/NewVariantEvalOutput/results/v1/tableexample.txt . Still need to address the issue where numeric columns are being interpreted as a vector of strings, not numbers.
kiran
2010-08-25 18:48:44 +0000
ac58eb3cbbSlightly better error message for the common error of only providing a dbsnp track but giving it zero clustering weight.
rpoplin
2010-08-25 18:41:21 +0000
5623e01602GenerateVariantClusters and VariantRecalibrator now uses hapmap and 1kg ROD bindings (in addition to dbsnp) to distinguish between knowns and novels. It no longer looks at by-hapmap validation status so providing hapmap is highly recommended. Example on the wiki. Input variants tracks now must start with input.
rpoplin
2010-08-25 18:33:40 +0000
bf0b6bd486Update integration tests to use the new ROD syntax.
hanna
2010-08-25 18:13:30 +0000
14198b74d5Can now compute av. qualities and stddevs per cycle for both original (when present in bam) and recalibrated quals
asivache
2010-08-25 17:14:58 +0000
23dbaa68e6Can design assays when multiple (distinct) events occur at the same locus (one assay per event)
asivache
2010-08-25 16:52:47 +0000
dc9e4098b2accidentally commited an old tool
aaron
2010-08-25 15:42:02 +0000
bc0826801cadding some utility code I've found helpful when working the Tribble index code
aaron
2010-08-25 15:38:47 +0000
4eff69d95eBack to using the LSF job name during dry runs since when the real job ids weren't available '-w(null)' wasn't too informative.
kshakir
2010-08-25 15:18:02 +0000
b4baa3eb8fCleanup. INDELS model is now disconnected (and renamed 'DINDEL' in preparation for adding plumbing for Guillermo soon)
ebanks
2010-08-25 14:52:51 +0000
3dc78855fdCommand-line argument tagging is in, and the ROD system is hacked slightly to support the new syntax (-B:name,type file) as well as the old syntax. Also, a bonus feature: BAMs can now be tagged at the command-line, which should allow us to get rid of some of the hackier calls in GenomeAnalysisEngine.
hanna
2010-08-25 03:47:57 +0000
85007ffa87Some clean up for the variant recalibrator. Now uses @Input and @Output so that it can join the Queue party. Users now specify a -o, -clusterFile, -tranchesFile, and -reportDatFile. Example on the wiki. ApplyVariantCuts now has an integration test. Base quality recalibrator now requires a dbsnp rod or vcf file. Now that the base quality recalibrator is using @Output the PrintStream shouldn't be closed in OnTraversalDone.
rpoplin
2010-08-24 20:14:58 +0000
f2b138d975Small refactoring: make Haplotype a public class since it will be soon extended and shared with other callers.
delangel
2010-08-24 17:52:36 +0000
43f1fb2380Okay, finally done with VCF compression. Now: 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
ebanks
2010-08-24 16:36:54 +0000
83624e3af3updated to deal with new cleaning pipeline outputs and potentially infinity TI/TV
corin
2010-08-24 16:01:09 +0000
25fb53e7a2Oops, forgot to call toLowerCase().
ebanks
2010-08-24 14:43:24 +0000
7957b60768We now automatically compress the output VCF if the file suffix is one of the supported types (.gz, .bz, .bz2). You can still specify -bzip if you want to use another file suffix (or pipe it to sdout for some reason).
ebanks
2010-08-24 14:39:59 +0000
7a8b6b87daCommitting Michael Yourshaw's patch for AnalyzeCovariates. We spawn each RScript process and wait for it to finish in series. Thanks Michael!
rpoplin
2010-08-24 13:06:25 +0000
44f3c5639aI have finally figured out that when you volunteer to do something in group meeting, you keep getting pestered about it on Mark's Omniplan doc until it gets done (except for contig aliasing, of course). As such...
ebanks
2010-08-24 04:14:50 +0000
691333f75cForce isRequired() to be false for @Deprecated args.
hanna
2010-08-23 23:50:30 +0000
5d6a6420a9New behavior for filling it output streams: if required==true for a field and the field 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.
hanna
2010-08-23 23:39:13 +0000
90aef66ec5Minor fixes for my last commit
ebanks
2010-08-23 23:25:29 +0000
ef795825fdYet more argument consistency updates
ebanks
2010-08-23 20:52:30 +0000
3aedd0055eUpdated firehose clean bam pipeline to pull firehose info and push back firehose clean bam.
kshakir
2010-08-23 20:38:42 +0000
7474afa7a3allow other objects access to the static method that resolves bam lists, and some renaming and improved documentation for the function.
aaron
2010-08-23 18:52:00 +0000
ccda4f6ec1More output consistency changes (updating wiki docs as I go along).
ebanks
2010-08-23 18:46:08 +0000
c9c6ff49c2Deprecated 'O' in favor of 'o' in the cleaner
ebanks
2010-08-23 18:09:24 +0000
55a8306a0dUpdate the @RMD tags to look for VariantContext.class instead of ReferenceOrderedDatum.class. Since the test for rod type is broken this won't affect anything right now.
ebanks
2010-08-23 17:49:37 +0000
35b9883dd6vcfwriter is in tribble now
aaron
2010-08-23 17:01:04 +0000
2d3b6d89dcadding the ability in Tribble to create indexes from a stream of features, so that we can create multiple indexes from one pass of the file. In the GATK we now create multiple indexes, and choose the most appropriate based on feature density, and the longest feature in the file. Also:
aaron
2010-08-23 06:54:59 +0000
295472bf69Simple change to handle a no-call (must avoid asking for the second allele, which will be be null in this case). Also, added a hack to deal with input VCFs where there are no genotype likelihoods (needed in order to process Hapmap and 1KG VCFs). In this mode, called genotypes are assigned a likelihood of 0.96, and alternative genotypes are given 0.02 each. I know Beagle actually takes genotype data without likelihoods, so this might not be the right way to do this.
kiran
2010-08-23 05:13:09 +0000
dec713a184Simple test code from Steve Schaffner to compute R^2 and D'. This is just for educational purposes. Don't use this code for anything, ever!
kiran
2010-08-23 05:06:16 +0000
8252494fa9Forgot to update UG performance test to reflect the new -o argument.
hanna
2010-08-23 00:57:16 +0000
c177801d81Add deprecated command-line arguments, and switched over UG to output to -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?"
hanna
2010-08-22 21:01:44 +0000
b80cf7d1d9Modifications to the output system for better interaction with @Output. Multiplexed arguments. More details in the Monday meeting.
hanna
2010-08-22 14:27:05 +0000
30a104228aDon't require entropy reduction when cleaning only at known sites; instead we need to trust the known indels. This will improve consistency between lane-level and aggregated cleaning.
ebanks
2010-08-22 02:44:38 +0000
b6989289fcPotential bug fix for bad references where some codons may have Ns
depristo
2010-08-21 12:09:33 +0000
121b4f23b6Simple change to allow a list of samples or regular expressions to be provided in a text file (one line per sample).
kiran
2010-08-21 00:01:48 +0000
165dc6d3b0Ryan, what did you decide about supporting this tool? Is it still useful?
ebanks
2010-08-20 19:16:14 +0000
2ef2f1b24aFix UG's simple indel calculation model so that deletions are created correctly
ebanks
2010-08-20 15:35:47 +0000
depristo
chartl
ebanks
aaron
fromer
kiran
rpoplin
delangel
hanna
kshakir
asivache
corin