7 Commits (406982284c906b0d1d9d71edc4e786c2eacfdc77)

Author	SHA1	Message	Date
Eric Banks	db2e0aaa1a	Darn, forgot to update unit tests.	2011-08-03 12:31:08 -04:00
Eric Banks	7c89fe01b3	Instead of having the padded reference base be some hackish attribute it is now an actual variable in the Variant Context class. More importantly, we now always require that it be present when padding is necessary - and validate as such upon construction of the VC. This cleans up the interface significantly because we no longer require that a reference base be passed in when writing a VC/VCF record.	2011-08-03 11:00:36 -04:00
Eric Banks	1afc49a297	There are some really 'interesting' (but apparently valid) records in the Mus musculus dbSNP file. Generalized the handling of complex cases in the dbSNP adaptor to handle it all. I just grabbed the actual Mus musculus dbSNP file as a test, ran it whole genome, and confirmed that we finally produce a valid VCF on it. Should be the last commit needed on this adaptor.	2011-07-28 13:55:58 -04:00
Eric Banks	6230315ff2	Along with my half-written commit message from earlier, I also forgot to commit the integration test updates. This is what happens when you try to do things 30 seconds before you leave for the day. To finish up from before: complex events weren't being padded with the reference base as per the VCF spec. They are now.	2011-07-27 22:51:21 -04:00
Eric Banks	a2597e7f00	This commit incorporates several different changes that each pretty much break all the VCF-based integration tests, so I bunched them all together. We now officially emit VCF4.1 files (woo hoo), which means that the VCF headers are now all different (header version is 4.1 plus counts for some of the annotations are 'A' or 'G'). Also, I've added a Read Filter for reads with MQ=255 ('unavailable' in the SAM spec) and have applied this to the UG and the RMS MQ annotation.	2011-07-12 14:11:53 -04:00
Eric Banks	33c67a139c	Wrong package; this should have been moved when VC got moved in from Tribble	2011-06-29 14:56:02 -04:00
David Roazen	3c9497788e	Reorganized the codebase beneath top-level public and private directories, ... removing the playground and oneoffprojects directories in the process. Updated build.xml accordingly.	2011-06-28 06:55:19 -04:00