CHANGES IN VERSION 1.10.0 (2014-10-01) -------------------------------------- NEW FEATURES o This release corresponds to C++ BitSeq version 0.7.5. o adding "unstranded" option to parseAlignment, to allow read pairs with various directions to be used o enable excluding singletons (single-mate alignments of paired reads) in parseAlignment CHANGES IN VERSION 1.5.5 (2013-10-12) ------------------------------------- BUG FIXES o fixed problem with processing Fasta reads o fixed problem with processing mixed alignments in empirical estimation o use proper includes for samtools CHANGES IN VERSION 1.5.2 (2013-10-09) ------------------------------------- BUG FIXES o fixed force option in estimateHyperPar (it worked the wrong way around) o fixed problem with effective lengths being too small NEW FEATURES o add estimateVBExpression function that uses Variational Bayes inference method to estimate expression levels of transcripts CHANGES IN VERSION 1.4.3 (2013-07-08) ------------------------------------- BUG FIXES o fixed major bug in getGeneExpression and getWithinGeneExpression that caused wrong results for some genes if transcripts were not grouped by genes o fixed bug in parseAlignment which would occasionally cause underflow in effective length computation o fixed taking square root of wrong column in getExpression result's $means field NEW FEATURES o much faster parseAlignment computation o improved precision in getGeneExpression and getWithinGeneExpression o getWithinGeneExpression provides option to keep transcripts in original order even if they are not grouped by genes CHANGES IN VERSION 1.3.11 (2013-04-02) ------------------------------------- NEW FEATURES o new functions for handling transcript information o getExpression returns also counts and transcript information CHANGES IN VERSION 1.3.10 (2013-03-30) ------------------------------------- NEW FEATURES o proper handling of alignments from Bowtie2 CHANGES IN VERSION 1.3.6 (2013-03-19) ------------------------------------- NEW FEATURES o enabling prallelization in getExpression via OpenMP o gene names can be extracted from Ensembl-like reference while estimating expression o added seed option for estimateHyperPar and estimateDE o improved output for getDE when using more than 2 conditions BUG FIXES o problem with long lines in reference sequence file in parseAlignment CHANGES IN VERSION 1.2.2 (2013-01-29) ------------------------------------- o occasional crash when using non-uniform read model in getExpression CHANGES IN VERSION 1.2.1 (2012-11-06) ------------------------------------- o parsing gapped alignments and half-alignments of paired end reads CHANGES IN VERSION 1.2.0 (2012-09-27) ------------------------------------- o IMPORTANT change: the way samples-files are passed to getDE, estimateHyperPar, estimateDE changed: instead of providing 2 vectors of filenames for each condition, the files are passed as a list of vectors, each vector containing filenames for one condition (allowing use of more than 2 conditions) o new internal structure (not visible to user) o estimateExpression has a new convergence criterion which should result in producing fewer samples. It is faster and avoids the use of MCMC_scaleReduction and MCMC_samplesNmax flags. o estimateExpressionLegacy uses the original convergence criterion o library normalization option for getDE, estimateDE, estimateHyperPar, getMeanVariance (in form of providing the normalization constants, for getting the constants please use edgeR or similar) CHANGES IN VERSION 1.0.1 (2012-04-30) ------------------------------------- o fixed bug in estimateExpression (reading file with probabilities occasionally failed when transcript information file was not specified) o improved manual, replaced overflowing lines o added option "pretend" to produce C++ calls CHANGES IN VERSION 1.0.0 (2012-04-01) ------------------------------------- o First Bioconductor release of BitSeq