USeq Source Code

DefinedRegionDifferentialSeq
* Changed the default minimum read count threshold to 20 from 10. This lower number was allowing too many low hit genes into the multiple testing correct and causing the FDRs to be significantly lower than necessary.

VCFComparator
* Added checks to convert 0|1 genotypes to 0/1, and 1/0 calls to 0/1; these are needed to standardize the calls and enable genotype matching with the NIST key

SamTranscriptomeParser
* Bug fix for merging paired alignments where an insertion occurred immediately before the start of the second read

Added modifications to support datasets lacking whole strands, needed for amplicon data.

Added modifications to support datasets lacking whole strands, needed for amplicon data.

Lots of mods to support the processing of methylated datasets where particular chromosomes only have + or - stranded data and not both. This breaks most everything. Ugg.

Added new beta app for annotating vcf variants with splice junction affects.

Bug Fix

Added ability to set the minimum base read coverage threshold, defaults to 5 alignments

Added support for David's best practices to TomatoFamer. See 'exome_best'

DefinedRegionDifferentialSeq
* Fixed bug when DESeq was outputting FDRs with "NA", these were being assigned the max FDR, switched it to 0.
Note, this won't effect most analysis since these genes also have a log2 of 0 and wouldn't survive standard log2Ratio thresholding.

NovoalignBisulfiteParser
* Fixed a bug in the -b4 novoalignment parsing
* Remove support for native novoalign format, just sam/ bam now.

Updates to bisulfite processing to support -b4 novoalignments, reinstatement of overlapping paired read collapsing, additions to VCF classes to allow for scoring splice junction effects.