c4dfc1fb8b
Temporary commit of parallelization support for RealignerTargetCreator. Tim begged us for this and I got assurances from Khalid/Matt that this would also be extremely helpful for the whole genome calling pipeline, so I spent a while working on this. Needs to be fixed up though because apparently only the leaves in the hierarchical reduce get their output aggregated. Worked out a better solution with Matt.
2011-10-06 13:41:36 -04:00
c3eff7451a
Found a small inefficiency while profiling: we were still using String.split instead of ParsingUtils.split to break up array values in the INFO field. There was a noticeable (albeit not big) difference in the change when reading sites only files.
2011-10-03 14:20:39 -04:00
c7898a9be7
inconsequential change in string constants printed into the vcf which noone uses anyway...
2011-09-30 16:40:21 -04:00
010899f886
Merge branch 'master' of ssh://gsa1/humgen/gsa-scr1/gsa-engineering/git/unstable
2011-09-30 15:51:09 -04:00
05fba6f23a
Clipping ends inside deletion and before insertion
...
fixed.
2011-09-30 15:44:43 -04:00
e0b771c233
Pre-sorting output
...
sorting variable regions independently to make the output pre-sorted. Major speedup.
2011-09-30 15:44:43 -04:00
b5bdea1cb9
Fix: hardclip first interval overlap but overlap subsequent intervals
...
If a read had the left tail clipped (that would overlap the current interval) but the right tail overlaps other intervals, this was triggering out of order reads to get processed before they should. Fixed.
2011-09-30 15:44:43 -04:00
d211b3b7f0
Fixing interval overlap logic
...
Complete re-write of the optimized interval overlap logic. Added lots of exceptions to trap future wrongdoings.
2011-09-30 15:44:42 -04:00
73577b72cf
Misc minor updates in HaplotypeCaller
2011-09-30 13:41:24 -04:00
af6c053435
Merge branch 'master' of ssh://nickel.broadinstitute.org/humgen/gsa-scr1/gsa-engineering/git/unstable
2011-09-30 13:33:31 -04:00
a881d6f145
Now only generates the poly VCF with select variants if the file doesn't exist
2011-09-30 08:42:09 -04:00
d901fed617
Merge branch 'master' of ssh://gsa1/humgen/gsa-scr1/gsa-engineering/git/unstable
2011-09-30 08:41:44 -04:00
cabacf028d
Intermediate commit to fix interval skipping
...
may need additional testing.
2011-09-29 18:45:12 -04:00
98ecaf8aa0
Support for ReducedReads with reduced counts and average quals
...
-- ReadUtils and UnitTest updated to support new byte[] style
-- Removed unnecessary read transformer in PairHMM
2011-09-29 17:18:39 -04:00
9508220157
fixed hard clipping both ends inside deletion
...
If both ends of the interval falls within a deletion in the read then hardClipBothEnds would cut the right tail first including the entire deletion, then fail to cut the left tail because there would not be any bases there anymore. Fixed.
2011-09-29 15:36:49 -04:00
a5e75cd14c
Outputting both consensus base qualities and counts
...
The base qualities of a consensus reads are now the average quality of the bases forming the consensus base (most common base) and the consensus quality tag now carry an array with the counts of each base in the consensus. This should increase file size but improve calling sensitivity/specificity.
2011-09-29 12:54:41 -04:00
d62f2f33bc
Added indel specific context size parameter
...
Parameter was added to the framework but implementing the functionality is pending.
2011-09-29 12:54:41 -04:00
21c4abdd36
Disabling all SlidingReadUnitTests
2011-09-29 12:20:35 -04:00
4086fa768f
Disabling all ReadClipperUnitTests
2011-09-29 12:20:35 -04:00
6dec932ca9
Merge branch 'master' of ssh://gsa3.broadinstitute.org/humgen/gsa-scr1/gsa-engineering/git/unstable
2011-09-29 11:47:13 -04:00
c08468eb9d
A couple of updates while trying to get desired R 2.13 compactPDF support.
...
preQC:
- For R 2.13 when parsing fingerprints explicitly coercing the text before parsing
- Added LOD geom_line() at +/-3 based on Tim's presentation at PM meeting (ppt to go to pipeline wiki asap)
- PF_INDEL_RATE of zero replaced with NA
- NA's are not "violations" auto filter samples since 0+NA = NA, and subset test only looks for 0 violations
- Restored plots for MEAN_READ_LENGTH, BAD_CYCLES, and MEDIAN_INSERT_SIZE by explicitly print()'ing the created plots
postQC:
- Fixed R 2.13 font scaling by moving size out of aes, except when using highlighting
- TODO: Don't know how to scale by aes for highlighting *and* use a smaller overall font size outside aes
2011-09-29 11:21:50 -04:00
e366ee18bc
Adding ability to read in and make use of kmer quality tables during HMM evaluation
2011-09-29 07:46:19 -04:00
9e32ede5b2
Merge branch 'master' of ssh://nickel.broadinstitute.org/humgen/gsa-scr1/gsa-engineering/git/unstable
2011-09-29 07:40:08 -04:00
fc86cd6fd8
Merge branch 'master' of ssh://nickel.broadinstitute.org/humgen/gsa-scr1/carneiro/gatk/RR into rr
2011-09-29 00:12:15 -04:00
4fd5630f6a
Added ReadClipper Unit Test
...
* Includes tests that include HardClip to Read and Reference Coords.
* Changed ReadUtils.HardClipByReferenceCoordinates from private to protected to allow for testing
2011-09-28 23:13:50 -04:00
f49a12de6b
Updating latest changes from the repository to reduce reads repo
2011-09-28 22:31:57 -04:00
9272ed03b5
Merged bug fix from Stable into Unstable
2011-09-28 21:26:43 -04:00
0acaf2df65
Fix an embarrassing issue where a specific configuration of minimal coverage
...
over small intervals could cause reads to be dropped from the pileup. Nothing
to see here...
2011-09-28 21:23:01 -04:00
07b0a75d96
Added SlidingRead Unit Test
...
Includes test clipStart and trimToVariableRegion
2011-09-28 21:22:57 -04:00
c5f1a4325f
Updated preQC:
...
- full 8.5x11
- concating multiple initiatives / bait_sets
- Using NA instead of python None when WR dates are unavailable
- In new aggregations where the sample may have per library metrics, only using the sample level metrics, i.e. library is null
Updated postQC:
- Renamed some variables to assist with traceback()
- Fixed crashes on batches with two alleles or two samples such as Seminara_MC_1_09222011 or Engle_MC_2_09222011
- Added dependency tracking to PostCallingQC.scala so that the R script does try to run before the evals are complete
Other minor cleanup.
Tried to use R 2.13 compactPDF but a few issues to work out with fingerprint boxplots in preQC and geom_text font size in postQC.
2011-09-28 20:23:30 -04:00
edf852d47d
Adding lists to ReduceReads script
...
script can handle single file or list of files separately now. Always scatter/gathering.
2011-09-28 18:40:30 -04:00
64e7b3000c
Fix read spans deletion through the entire interval
...
if the read has a deletion that spans the entire length of the interval, it should not be added to mapped reads.
2011-09-28 18:40:30 -04:00
a93ece07e3
ScatterGatherable reduce reads script
...
Get your reduce read in a matter of seconds...
2011-09-28 18:40:30 -04:00
c8d3a720f9
Merge branch 'master' of ssh://nickel.broadinstitute.org/humgen/gsa-scr1/gsa-engineering/git/unstable
2011-09-28 18:17:34 -04:00
7e3cb45093
Further performance optim in banded hmm, about 60% speed improvement over current implementation now
2011-09-28 16:27:28 -04:00
1b1ca80df2
Merge branch 'master' of ssh://nickel.broadinstitute.org/humgen/gsa-scr1/gsa-engineering/git/unstable
2011-09-28 16:17:39 -04:00
3b73dc89fe
Making several esoteric arguments in the BQSR @Hidden. Adding basic support for Complete Genomics machine cycle.
2011-09-28 16:17:31 -04:00
ff2f4df043
Fixed hardclipping inside indel (right tail)
...
when hard clipping the right tail of a read falls inside a deletion, clipping should fall back to the last base before the deletion to follow the ReadClipper's contract.
2011-09-28 16:07:34 -04:00
3c7b7f74ef
Optimized interval iteration
...
Using a TreedSet to manipulate getToolkit.getIntervals() and being smart about which intervals to test makes interval clipping O(1) instead of O(n).
2011-09-28 16:07:34 -04:00
5c9b659c02
clipping both ends of the reads was modifying the original read
...
This goes against the ReadClipper contract, and was affecting the second part of the read that spans over multiple intervals. Fixed.
2011-09-28 16:07:34 -04:00
fe23e4d10c
Merge branch 'master' of ssh://nickel.broadinstitute.org/humgen/gsa-scr1/gsa-engineering/git/unstable
2011-09-28 15:53:11 -04:00
e2b9030e93
First mostly fully functional implementation of banded pair HMM likelihood computation for indel caller. More experimentation to follow but it right now works in small data sets and at least it doesn't break existing things. Disabled by default at this point
2011-09-28 15:51:48 -04:00
1b45f21774
Removing this command-line tool. Purposely not doing this in stable so that users who may still use it have time to find other options. But the docs are no longer on the wiki.
2011-09-28 13:18:32 -04:00
1f0e354fae
Merge branch 'master' of ssh://nickel.broadinstitute.org/humgen/gsa-scr1/gsa-engineering/git/unstable
2011-09-28 13:13:21 -04:00
bb619a9a3c
Fixing docs
2011-09-28 13:13:03 -04:00
5812004e06
Merge branch 'stable'
2011-09-28 11:36:40 -04:00
a88b7c1203
Merge branch 'master' of ssh://gsa1/humgen/gsa-scr1/gsa-engineering/git/unstable
2011-09-28 11:36:33 -04:00
a5006831d7
Shows "" not empty space when default string value is ""
2011-09-28 11:35:52 -04:00
1e32281a15
Fix to not show -null when missing short name argument
2011-09-28 11:31:20 -04:00
fd287da189
Final tweaking of Smith-Waterman parameters based on large-scale parameter search with Queue
2011-09-28 11:22:01 -04:00
Eric Banks
Andrey Sivachenko
Mark DePristo
Mauricio Carneiro
Ryan Poplin
Khalid Shakir
Roger Zurawicki
Matt Hanna
Guillermo del Angel