7a0f6feda4
Make sure that too many alternate alleles aren't being passed to the genotyper (10 for now) and exit with a UserError if there are.
2011-12-05 16:18:52 -05:00
7fac4afab3
Fixed priors (now initialized upon engine startup in a multi-dimensional array) and cell coefficients (properly handles the generalized closed form representation for multiple alleles).
2011-12-05 15:57:25 -05:00
a7cb941417
The posteriors vector is now 2 dimensional so that it supports multiple alleles (although the UG is still hard-coded to use only array[0] for now); the exact model now collapses probabilities for all conformations over a given AC into the posteriors array (in the appropriate dimension). Fixed a bug where the priors and posteriors were being passed in swapped.
2011-12-04 13:02:53 -05:00
eab2b76c9b
Added loads of comments for future reference
2011-12-03 23:54:42 -05:00
29662be3d7
Fixed bug where k=2N case wasn't properly being computed. Added optimization for BB genotype case not in old model. At this point, integration tests pass except for 1 case where QUALs differ by 0.01 (this is okay because I occasionally need to compute extra cells in the matrix which affects the approximations) and 2 cases where multi-allelic indels are being genotyped (some work still needs to be done to support them).
2011-12-03 23:12:04 -05:00
71f793b71b
First partially working version of the multi-allelic version of the Exact AF calculation
2011-12-02 14:13:14 -05:00
91413cf0d9
Merged bug fix from Stable into Unstable
2011-11-29 14:01:23 -05:00
cb284eebde
Further updating VQSR tutorial wiki docs to reflect the bundle
2011-11-29 14:00:57 -05:00
dcb889665d
Merged bug fix from Stable into Unstable
2011-11-29 09:58:49 -05:00
447e9bff9e
Updating VQSR tutorial wiki docs to reflect the bundle
2011-11-29 09:57:45 -05:00
110298322c
Adding Transmission Disequilibrium Test annotation to VariantAnnotator and integration test to test it.
2011-11-29 09:29:18 -05:00
d7d8b8e380
Tribble v42 changes the Codec.canDecode method to take in a String instead of a File; this is something that Jim was adamant about (because Tribble can handle streams other than files). I didn't want the next person who needed to rev Tribble to deal with this change additionally, so I took care of updating the GATK now.
2011-11-28 14:18:28 -05:00
436b4dc855
Updated docs
2011-11-28 08:59:48 -05:00
e60272975a
Fix for changed MD5 in streaming VCF test
2011-11-23 19:01:33 -05:00
12f09d88f9
Removing references to SimpleMetricsByAC
2011-11-23 16:08:18 -05:00
fdd90825a1
Queue now outputs a GATK-like header with version number, build timestamp, etc.
2011-11-23 14:28:35 -05:00
e319079c32
Merge branch 'master' of ssh://gsa1/humgen/gsa-scr1/gsa-engineering/git/unstable
2011-11-23 13:02:11 -05:00
4107636144
VariantEval updates
...
-- Performance optimizations
-- Tables now are cleanly formatted (floats are %.2f printed)
-- VariantSummary is a standard report now
-- Removed CompEvalGenotypes (it didn't do anything)
-- Deleted unused classes in GenotypeConcordance
-- Updates integration tests as appropriate
2011-11-23 13:02:07 -05:00
e5b85f0a78
A toString() method for IntervalBindings
...
Necessary since we're currently writing things like this to our VCF headers:
intervals=[org.broadinstitute.sting.commandline.IntervalBinding@4ce66f56]
2011-11-23 11:56:12 -05:00
5a4856b82e
GATKReports now support a format field per column
...
-- You can tell the table to format your object with "%.2f" for example.
2011-11-23 11:31:04 -05:00
c8bf7d2099
Check for null comment
2011-11-23 10:47:21 -05:00
c1ea53d088
Solve merge conflicts
2011-11-23 09:17:32 -05:00
d2499bcc33
Cleaning up and institutionalizing several local hacks to ValidationSiteSelector: add argument to ignore polymorphic status in VC (useful when we want to intentionally select monomorphic sites), add dummy NullSampleSelector class that should be used if we really don't want to do any sample selection, and enabled sampleMode=NONE for this purpose.
2011-11-23 09:14:10 -05:00
6c2555885c
Caching getSimpleName() in VariantEval is a big performance improvement
...
-- Removed the SimpleMetricsByAC table, as one should just use the AlleleCount Stratefication and the upcoming VariantSummary table
2011-11-23 08:34:05 -05:00
32adbd614f
Solve merge conflict
2011-11-22 22:48:46 -05:00
941f3784dc
Solve merge conflict
2011-11-22 22:48:03 -05:00
75d93e6335
Another corner condition fix: skip likelihood computation in case we cut so many bases there's no haplotype or read left
2011-11-22 22:46:12 -05:00
a3aef8fa53
Final performance optimization for GenotypesContext
2011-11-22 17:19:30 -05:00
990c02e4de
Merge branch 'master' of ssh://gsa1/humgen/gsa-scr1/gsa-engineering/git/unstable
2011-11-22 17:19:11 -05:00
0fa50056eb
Reorganizing package structure for compression tools (incl. Reduce Reads)
...
ReduceReads is no longer the only compression tool. Reorganizing the package structure to reflect that.
2011-11-22 15:03:31 -05:00
1a50d54c03
A walker that generates the distribution of quality scores
...
Outputs a GATKReportTable that will be used as input to the Quantization walker. Eventually this functionality may be merged into ReduceReads or CountCovariates to avoid another traversal.
2011-11-22 14:49:28 -05:00
1614ca1115
Force use of LinkedList
...
Disambiguating which collection I need to use for mapped reads.
2011-11-22 14:49:28 -05:00
38a90da92c
Fixed merge conflict to Unstable
2011-11-22 14:39:45 -05:00
32a77a8a56
Prevent out of bound error in case read span > reference context + indel length. Can happen in RNAseq reads with long N CIGAR operators in the middle.
2011-11-22 13:57:24 -05:00
5821c11fad
For BAM and Reviewed errors we now check the error message to see if it's actually a 'too many open files' problem and, if so, we generate a User Error instead.
2011-11-22 10:50:22 -05:00
5e7b9ae119
The NONE option is not supported yet (but the error message should mention this)
2011-11-22 10:48:38 -05:00
7087310373
Embarassing bug fixed
2011-11-22 10:16:36 -05:00
e484625594
GenotypesContext now updates cached data for add, set, replace operations when possible
...
-- Involved separately managing the sample -> offset and sample sorted list operations. This should improve performance throughout the system
2011-11-22 08:40:48 -05:00
29ca24694a
UG now encoding NO_CALLs as ./. not ./.:.:4:0,0,0
...
A few updated UGs integration tests
2011-11-22 08:22:32 -05:00
4ea74c42de
Including GENCODE BED argument, removing CAPTURE target
2011-11-21 19:18:43 -05:00
2b51c01df4
Merge branch 'master' of ssh://gsa1/humgen/gsa-scr1/gsa-engineering/git/unstable
2011-11-21 19:16:06 -05:00
5443d3634a
Again, fixing the add call when we really mean replace
...
-- Updating MD5s for UG to reflect that what was previously called ./.:.:10:0,0,0 is now just ./. Eric will fix long-standing bug in QD observed from this change
-- VFW MD5s restored to their old correct values. There was a bug in my implementation to caused the genotypes to not be parsed from the lazy output even through the header was incorrect.
2011-11-21 19:15:56 -05:00
8c7430e6ff
Bugfix: isOriginalRead had the wrong check for WGS
...
The original read has to be contained in the current interval and the first read in the list. In WGS all reads are 'original'.
2011-11-21 17:12:11 -05:00
756695ee3f
Better Unit tests for Synthetic Reads
2011-11-21 17:12:11 -05:00
5ad3dfcd62
BugFix: byte overflow in SyntheticRead compressed base counts
...
* fixed and added unit test
2011-11-21 17:11:50 -05:00
9ea7b70a02
Added decode method to LazyGenotypesContext
...
-- AbstractVCFCodec calls this if the samples are not sorted. Previously called getGenotypes() which didn't actually trigger the decode
2011-11-21 16:21:23 -05:00
ab2efe3bd3
Reverting bad exact model changes
2011-11-21 16:14:40 -05:00
44554b2bfd
Merge branch 'master' of ssh://nickel.broadinstitute.org/humgen/gsa-scr1/gsa-engineering/git/unstable
2011-11-21 15:01:45 -05:00
022832bd74
Very bad use of the == operator with Strings was ensuring that validating GenomeLocs was very inefficient. This fix resulted in a significant speedup for a simple RodWalker.
2011-11-21 14:49:47 -05:00
1561af22af
Exact model code cleanup
...
-- Fixed up code when fixing a bug detected by aggressive contracts in GenotypesContext.
2011-11-21 14:35:15 -05:00
Eric Banks
Ryan Poplin
Mark DePristo
David Roazen
Guillermo del Angel
Mauricio Carneiro