311dfa0998
Now builds examples, as I expected. GATKPaperGenotyper lives again.
...
git-svn-id: file:///humgen/gsa-scr1/gsa-engineering/svn_contents/trunk@5953 348d0f76-0448-11de-a6fe-93d51630548a
2011-06-07 00:13:44 +00:00
2901abf070
Switch from PriorityQueue to TreeSet for better and more consistent performance.
...
git-svn-id: file:///humgen/gsa-scr1/gsa-engineering/svn_contents/trunk@5952 348d0f76-0448-11de-a6fe-93d51630548a
2011-06-06 20:41:30 +00:00
2c57721ed2
Updated printouts to help with debugging. Issue does appear to be deterministic though.
...
git-svn-id: file:///humgen/gsa-scr1/gsa-engineering/svn_contents/trunk@5950 348d0f76-0448-11de-a6fe-93d51630548a
2011-06-06 01:04:07 +00:00
27dfb53d26
We really don't want to be advising the user to use an unsafe option - really, they should fix their busted bam file.
...
git-svn-id: file:///humgen/gsa-scr1/gsa-engineering/svn_contents/trunk@5949 348d0f76-0448-11de-a6fe-93d51630548a
2011-06-05 05:18:16 +00:00
7e49e1668f
Finished changing md5's due to recent change in definition of mixed and indel vc's.
...
git-svn-id: file:///humgen/gsa-scr1/gsa-engineering/svn_contents/trunk@5948 348d0f76-0448-11de-a6fe-93d51630548a
2011-06-05 00:40:51 +00:00
d534241f35
Major revamp of annotations for indels:
...
a) All rank sum tests now work for indels including multiallelic sites. For the latter cases, rank sum test is REF vs most common allele
b) Redid computation of HaplotypeScore for indels. It's now trivially easy to do because we are already computing likelihoods of each read vs haplotypes in GL computation so we reuse that if available. For multiallelic case, we score against N haplotypes where N is total called alleles.
Drawback is that all cases need information contained in likelihood table that stores likelihood for each pileup element, for each allele. If this table is not available we dont annotate, so we can only fully annotate indels right now when running UG but not when running VariantAnnotator alone.
git-svn-id: file:///humgen/gsa-scr1/gsa-engineering/svn_contents/trunk@5947 348d0f76-0448-11de-a6fe-93d51630548a
2011-06-04 15:34:24 +00:00
1448a1f155
Change md5 because conversion of a tri-allelic dbsnp indel record is now legit
...
git-svn-id: file:///humgen/gsa-scr1/gsa-engineering/svn_contents/trunk@5946 348d0f76-0448-11de-a6fe-93d51630548a
2011-06-04 11:24:16 +00:00
53667ce8fa
Disabled test that checks whether output is the same whether in Genotype Given Alleles mode or not - it won't as long as extended events are finally removed.
...
git-svn-id: file:///humgen/gsa-scr1/gsa-engineering/svn_contents/trunk@5945 348d0f76-0448-11de-a6fe-93d51630548a
2011-06-04 00:52:54 +00:00
35df80de14
Updated md5 due to changes to changes in QUAL field when in Genotype given alleles mode w/indels when in insertions.
...
git-svn-id: file:///humgen/gsa-scr1/gsa-engineering/svn_contents/trunk@5944 348d0f76-0448-11de-a6fe-93d51630548a
2011-06-03 23:52:38 +00:00
b93829e505
The underlying bam file for this test was busted for many reasons preventing Picard folks from making unrelated changes, so I needed to fix it. Updating md5s accordingly.
...
git-svn-id: file:///humgen/gsa-scr1/gsa-engineering/svn_contents/trunk@5943 348d0f76-0448-11de-a6fe-93d51630548a
2011-06-03 22:26:06 +00:00
a8faacda4e
Major change to UG engine to support:
...
a) Genotype given alleles with indels
b) Genotyping and computing likelihoods of multi-allelic sites.
When GGA option is enabled, indels will be called on regular pileups, not on extended pileups (extended pileups will be removed shortly in a next iteration). As a result, likelihood computation is suboptimal since we can't see reads that start with an insertion right after a position, and hence quality of some insertions is removed and we could be missing a few marginal calls, but it makes everything else much simpler.
For multiallelic sites, we currently can't call them in discovery mode but we can genotype them and compute/report full PL's on them (annotation support comes in next commit). There are several suboptimal approximations made in exact model to compute this. Ideally, joint likelihood Pr(Data | AC1=i,AC2=j..) should be computed but this is hard. Instead, marginal likelihoods are computed Pr(Data | ACi=k) for all i,k, and QUAL is based on highest likelihood allele.
git-svn-id: file:///humgen/gsa-scr1/gsa-engineering/svn_contents/trunk@5941 348d0f76-0448-11de-a6fe-93d51630548a
2011-06-03 22:13:07 +00:00
cd293f145b
More stable reduced reads representation. Bug fixes throughout. No diffs by <1% of sites in an exome, and the majority of these differences are filtered out, or are obvious artifacts. UnitTests for BaseCounts. BaseCounts extended to handle indels, but not yet enabled in the consensus reads.
...
git-svn-id: file:///humgen/gsa-scr1/gsa-engineering/svn_contents/trunk@5939 348d0f76-0448-11de-a6fe-93d51630548a
2011-06-03 20:11:31 +00:00
80cbc1924b
Oops, just realized that I forgot to comment my commit from yesterday so it was clear what was happening
...
git-svn-id: file:///humgen/gsa-scr1/gsa-engineering/svn_contents/trunk@5938 348d0f76-0448-11de-a6fe-93d51630548a
2011-06-03 18:06:41 +00:00
e4eb8087bc
A VariantContext can now be isSymbolic. More importantly, multi-allelic variants are now properly handled in determining their type [using isMixed only if any of the biallelic variant types differ between the alt alleles].
...
git-svn-id: file:///humgen/gsa-scr1/gsa-engineering/svn_contents/trunk@5937 348d0f76-0448-11de-a6fe-93d51630548a
2011-06-03 18:02:47 +00:00
b4c479bcb0
Support for reducedReads in the pileup and UG. Totally experimental -- the code interface could change, and so could the implementation. Only works for SNPs now. Pileup has contracts as well.
...
git-svn-id: file:///humgen/gsa-scr1/gsa-engineering/svn_contents/trunk@5936 348d0f76-0448-11de-a6fe-93d51630548a
2011-06-03 16:39:01 +00:00
2df12472c2
One more step in commit to support multi-allelic indel genotyping and processing: utility class that supports multi-allelic genotype likelihoods
...
git-svn-id: file:///humgen/gsa-scr1/gsa-engineering/svn_contents/trunk@5935 348d0f76-0448-11de-a6fe-93d51630548a
2011-06-03 16:08:29 +00:00
420d8feff6
No one should be calling the createHeader method(s) directly, but instead should be going through the full readHeader method (because it first sets the version); therefore I made them package protected and merged them. Updated the various unit tests that were using createHeader and were dangerously assuming that the header version was defaulting to 4.0 (which it no longer does).
...
git-svn-id: file:///humgen/gsa-scr1/gsa-engineering/svn_contents/trunk@5934 348d0f76-0448-11de-a6fe-93d51630548a
2011-06-03 02:17:37 +00:00
86df10ec09
UnitTests for ConsensusSpan infrastructure
...
git-svn-id: file:///humgen/gsa-scr1/gsa-engineering/svn_contents/trunk@5929 348d0f76-0448-11de-a6fe-93d51630548a
2011-06-02 22:44:52 +00:00
74298f6858
Basic walker to calculate statistics of CNV genotyping copy counts
...
git-svn-id: file:///humgen/gsa-scr1/gsa-engineering/svn_contents/trunk@5927 348d0f76-0448-11de-a6fe-93d51630548a
2011-06-02 21:46:35 +00:00
ad9dca9137
Package updated. Copyrights added
...
git-svn-id: file:///humgen/gsa-scr1/gsa-engineering/svn_contents/trunk@5926 348d0f76-0448-11de-a6fe-93d51630548a
2011-06-02 21:29:27 +00:00
3d628f06f0
moved to playground
...
git-svn-id: file:///humgen/gsa-scr1/gsa-engineering/svn_contents/trunk@5925 348d0f76-0448-11de-a6fe-93d51630548a
2011-06-02 21:25:26 +00:00
429833c05a
Intermediate commit (DVCS, where are you?) of a fully operational ReducedRead walker. Now results in minor differences in the raw calls (filtering is a different matter) in an exome but 20x less disk space than the full exome data. Changes to the UG necessary to process reduced reads are not yet committed, as they are being tested. This code is being moved to playground now.
...
git-svn-id: file:///humgen/gsa-scr1/gsa-engineering/svn_contents/trunk@5924 348d0f76-0448-11de-a6fe-93d51630548a
2011-06-02 21:13:31 +00:00
dd6d61c031
Adding integration test to cover the case of a read that only covers an insertion (i.e. no M in the CIGAR string).
...
git-svn-id: file:///humgen/gsa-scr1/gsa-engineering/svn_contents/trunk@5923 348d0f76-0448-11de-a6fe-93d51630548a
2011-06-02 21:02:47 +00:00
d0ca6f8a9c
Patch for case that a read spans only an insertion (i.e. no Ms in the CIGAR string): the end position should not be less than the start position (which is how Picard defines it) but instead should be equal to it. This is just a patch; we'll get a proper solution in at some point.
...
git-svn-id: file:///humgen/gsa-scr1/gsa-engineering/svn_contents/trunk@5922 348d0f76-0448-11de-a6fe-93d51630548a
2011-06-02 20:40:56 +00:00
3302a733ef
Fixed docs
...
git-svn-id: file:///humgen/gsa-scr1/gsa-engineering/svn_contents/trunk@5920 348d0f76-0448-11de-a6fe-93d51630548a
2011-06-02 16:02:14 +00:00
84c2c5d7e6
Stop running away from my commits, test modules.
...
git-svn-id: file:///humgen/gsa-scr1/gsa-engineering/svn_contents/trunk@5919 348d0f76-0448-11de-a6fe-93d51630548a
2011-06-02 13:05:53 +00:00
092952db44
After verifying that the changes to these tests were all in the RankSum annotations, I'm commiting fixes to the test md5s.
...
git-svn-id: file:///humgen/gsa-scr1/gsa-engineering/svn_contents/trunk@5918 348d0f76-0448-11de-a6fe-93d51630548a
2011-06-02 13:01:18 +00:00
c7fe062cb7
Refactored the VCF codec classes to minimize code duplication (which happened during the VCF3/4 split). Now, both codecs extend the AbstractVCFCodec class and all shared functionality exists there. Only methods that differ between the various codecs (e.g. because FILTER strings are encoded differently) are defined in the actual codecs. While I was in there, I put in checks for invalid empty inputs in the ID, FILTER, and INFO fields.
...
git-svn-id: file:///humgen/gsa-scr1/gsa-engineering/svn_contents/trunk@5917 348d0f76-0448-11de-a6fe-93d51630548a
2011-06-01 19:40:47 +00:00
81d9808eea
Next version of test output for non-determinism
...
git-svn-id: file:///humgen/gsa-scr1/gsa-engineering/svn_contents/trunk@5916 348d0f76-0448-11de-a6fe-93d51630548a
2011-06-01 19:36:56 +00:00
511cd48d7a
There is an edge case ( |Set1| = 5, |Set2| = 4) where the exact p-value exceeds the range of the normal distribution we want to invert. For the edge cases, this happens exactly at the mean, and so this can be safely replaced with a z value of 0.0
...
git-svn-id: file:///humgen/gsa-scr1/gsa-engineering/svn_contents/trunk@5915 348d0f76-0448-11de-a6fe-93d51630548a
2011-06-01 17:30:09 +00:00
dcd13060e1
created wiki page for Print Reads and changed help to match wiki.
...
git-svn-id: file:///humgen/gsa-scr1/gsa-engineering/svn_contents/trunk@5914 348d0f76-0448-11de-a6fe-93d51630548a
2011-06-01 16:26:32 +00:00
8f6af299d8
Remove what is hopefully the last of the evil core -> playground dependencies.
...
git-svn-id: file:///humgen/gsa-scr1/gsa-engineering/svn_contents/trunk@5913 348d0f76-0448-11de-a6fe-93d51630548a
2011-06-01 16:22:35 +00:00
8f3e8f934d
added a quick option to print the first n reads.
...
git-svn-id: file:///humgen/gsa-scr1/gsa-engineering/svn_contents/trunk@5912 348d0f76-0448-11de-a6fe-93d51630548a
2011-06-01 16:16:50 +00:00
a79967d9af
After extensive testing of MannWhitneyU:
...
- Verified that exact calculations do agree with R's dwilcox()
- Verified that exact calculations do not agree with R's wilcox.test
+ This is because R does a correction, and calculates CDFs rather than PDFs (e.g. sums over dwilcox() values)
- Can now specify MWU to calculate cumulative exact tests, rather than point probabilities
- Z-scores are now calculated properly for exact tests
+ Previously, z-values calculated by inverting normal CDF from U-statistic PDF
+ Now both inversions are done, with a smart heuristic (biased variance) to make the point-calculated Z-value more accurate
+ Additional tests
git-svn-id: file:///humgen/gsa-scr1/gsa-engineering/svn_contents/trunk@5911 348d0f76-0448-11de-a6fe-93d51630548a
2011-06-01 15:51:27 +00:00
2b5683909e
Updated VQSR integration tests because of the new Omni file. Fixed overflow condition in FisherStrand when the depth is too high.
...
git-svn-id: file:///humgen/gsa-scr1/gsa-engineering/svn_contents/trunk@5910 348d0f76-0448-11de-a6fe-93d51630548a
2011-06-01 14:20:37 +00:00
6cc84c3ce2
Make the set of VariantContextAdaptors dynamic so that Andrey's MafFeature can
...
continue to exist and live in playground (and thus outside of the normal release
/ git release branch).
git-svn-id: file:///humgen/gsa-scr1/gsa-engineering/svn_contents/trunk@5909 348d0f76-0448-11de-a6fe-93d51630548a
2011-06-01 02:54:55 +00:00
44cb7e4980
Renaming to make grepping through the output less confusing
...
git-svn-id: file:///humgen/gsa-scr1/gsa-engineering/svn_contents/trunk@5908 348d0f76-0448-11de-a6fe-93d51630548a
2011-05-31 19:54:44 +00:00
b75583a90b
Adding debug statements for David to aid in testing the non-determinism problem. I wouldn't recommend running with --stats temporarily (or ever in fact, which is why it's @Hidden).
...
git-svn-id: file:///humgen/gsa-scr1/gsa-engineering/svn_contents/trunk@5907 348d0f76-0448-11de-a6fe-93d51630548a
2011-05-31 19:53:59 +00:00
c50d290133
Removing printf's used for debugging -- they have served their purpose.
...
git-svn-id: file:///humgen/gsa-scr1/gsa-engineering/svn_contents/trunk@5906 348d0f76-0448-11de-a6fe-93d51630548a
2011-05-31 14:06:37 +00:00
0aef5c0074
Totaly experimental, possibly useless annotation that logs # of MQ0 reads / total depth, TBD if VQSR can use it.
...
git-svn-id: file:///humgen/gsa-scr1/gsa-engineering/svn_contents/trunk@5905 348d0f76-0448-11de-a6fe-93d51630548a
2011-05-30 14:05:39 +00:00
b4d379584c
Commented out the generation of the GATKReport that I was using for debugging.
...
git-svn-id: file:///humgen/gsa-scr1/gsa-engineering/svn_contents/trunk@5903 348d0f76-0448-11de-a6fe-93d51630548a
2011-05-27 22:15:09 +00:00
2a9c75c5ba
Throw an exception if the programmer tries to access a column that doesn't exist.
...
git-svn-id: file:///humgen/gsa-scr1/gsa-engineering/svn_contents/trunk@5902 348d0f76-0448-11de-a6fe-93d51630548a
2011-05-27 22:08:48 +00:00
f3b38c0d3e
Fixed a bug in my math where I assumed the genotype likelihoods were normalized to 1.0 when they in fact are not. *Now* genotypes get altered when a different genotype configuration leads to a more consistent answer with regards to inheritance constraints. There's the question of what to do when two configurations are almost equally likely - I should probably filter those events out. But currently there is no threshold on the transmission probability.
...
git-svn-id: file:///humgen/gsa-scr1/gsa-engineering/svn_contents/trunk@5901 348d0f76-0448-11de-a6fe-93d51630548a
2011-05-27 22:08:05 +00:00
5974675b43
Two intermediate commits, to work over the weekend.
...
ReplicationValidationWalker: Just the skeleton of what will be the implementation of the replication/validation model.
dataProcessingV2: Committing an UNTESTED implementation of BWA alignment. I am running tests on it over the weekend.
git-svn-id: file:///humgen/gsa-scr1/gsa-engineering/svn_contents/trunk@5900 348d0f76-0448-11de-a6fe-93d51630548a
2011-05-27 22:03:08 +00:00
69d9b5989f
documenting this walker as it may be useful to others in the future.
...
git-svn-id: file:///humgen/gsa-scr1/gsa-engineering/svn_contents/trunk@5899 348d0f76-0448-11de-a6fe-93d51630548a
2011-05-27 21:58:51 +00:00
a50c40ed05
Temporary commit to aid in investigation of recent intermittent
...
IndelRealignerIntegrationTest failures -- yes, it's the classic printf()
debugging technique. Will revert in a day or two once I get the data I need :)
git-svn-id: file:///humgen/gsa-scr1/gsa-engineering/svn_contents/trunk@5896 348d0f76-0448-11de-a6fe-93d51630548a
2011-05-27 20:01:57 +00:00
2227f49220
misc cleanup
...
git-svn-id: file:///humgen/gsa-scr1/gsa-engineering/svn_contents/trunk@5893 348d0f76-0448-11de-a6fe-93d51630548a
2011-05-27 16:49:20 +00:00
9e834391fe
We now skip over all covering RODs in the BQSR as intended instead of just those which can be converted into a VariantContext. All the integration tests change because of subtleties in how certain dbsnp rod records are being converted into VCs. Added integration test which uses a bed file as the list of known polymorphic sites.
...
git-svn-id: file:///humgen/gsa-scr1/gsa-engineering/svn_contents/trunk@5892 348d0f76-0448-11de-a6fe-93d51630548a
2011-05-27 16:32:17 +00:00
8ed82e5a08
The previous version of the UG was always creating BAQ'd pileups for the underlying site QUAL calculation. This resulted in some slowdown in the code. But as far as I can tell, the code actually didn't apply the BAQ'd base quality anywhere when the BAQ field wasn't in the read, so this just saves us 20% of the runtime when BAQ isn't enabled from heading into the BAQ subsystem when we don't actually want to get the BAQ'd base qualities.
...
Fixed minor problem with WalkerTest for "" (for parameterization) md5s.
Added an explicit integrationtest for BAQ NONE
Now only creates the BAQ'd pileup, if the useBAQPileup parameter is provide in initializeAlternateAllele.
git-svn-id: file:///humgen/gsa-scr1/gsa-engineering/svn_contents/trunk@5891 348d0f76-0448-11de-a6fe-93d51630548a
2011-05-27 14:00:52 +00:00
136c8c7900
ClipReads now supports HARDCLIP_BASES, though in fact this turned out to be not necessary for my desired tests. In the process of developing the HARDCLIP mode, I added some proper ReadUtils unit tests, which would ideally be expanded to include other ReadUtil functions, as added
...
git-svn-id: file:///humgen/gsa-scr1/gsa-engineering/svn_contents/trunk@5890 348d0f76-0448-11de-a6fe-93d51630548a
2011-05-27 11:42:22 +00:00
depristo
alecw
ebanks
delangel
fromer
chartl
carneiro
droazen
rpoplin
hanna
kiran