Version 1.7.1 - Added findPeakHeight function to estimate optimal minHeight value based on False Discovery Rate for using in enrichedPeaks Version 1.5.1 - Fixed bug in enrichedRegions method for RangedDataList objects VERSION 1.3.1 - Fixed bug in enrichedRegions that caused it to crash when a region had 0 counts in all samples - Fixed bug in plotMeanCoverage VERSION 1.1.5 - Added missing documentation VERSION 1.1.4 - Fixed bug in cmds for method with empty chromosomes VERSION 1.1.3 - Fixed bug in alignPeaks (no longer directly accessing matchMatrix slot of RangesMatchingList object) - Added plotMeanCoverage - Fixed bug in gini for method with no chromosome lengths VERSION 1.1.2 - Add methods for objects of class GRanges and GRangesList. - Add set.seed to vignette. - Use pvec instead of mclapply in tabDuplReads and filterDuplReads. - Correct rpkm form enrichedRegions and remove it from enrichedPeaks - Fixed giniCoverage for ranges with low numbers of reads VERSION 1.1.1 - Adjusted parallel computing in enrichedPeaks so that it no longer spans an uncontrolled number of child processes when mc.cores>1 - Added arguments "labels" and "cex.text" to the plot method for cmdsFit objects. - Added monotonicity contraint to filterDuplReads and fdrEnrichedCounts to ensure that the estimated FDR decreases with the number of repeats - Fixed overflow problem in enrichedChrRegions which occurred for long genomes (e.g. human) - Adjusted behavior of stdPeakLocation so that it is consistent with PeakLocation - Fixed bug in RPKM calculation by enrichedRegions - Added option to compute Spearman correlations in cmds VERSION 0.99 - Adapted enrichedChrRegions to adapt to change in IRanges behavior - Fixed bug in PeakLocation and stdPeakLocation when strand information was stored in variables of type factor - Included adjustments for number of reads in each sample for giniCoverage and ssdCoverage - Improved filterDuplReads, using mixtures of NegBinom - Adjusted behavior of giniCoverage - Removed readAlignedBatch, export2Aligned, export2SGR, export2SAM VERSION 0.2.7 - Fixed bug in filterDuplReads method for 'RangedDataList' which caused it not to pass on arguments to 'RangedData' method - Changed default value of ppOverAmp=.95 to .99 in filterDuplReads - Added function rowLogRegLRT - Fixed bug in alignPeaks caused by change in IRanges behavior. - Added function coverageDiff to compute difference in coverage between two objects - Fixed bug regionsCoverage caused by convertion of class "by" objects - Added plotChrRegions function - Added enrichedChrRegions function - Added countHitsWindow function - stdPeakLocation now allows to set xlim - plot method for cmdsFit objects now automatically puts both axis on the same scale - closestGene now allows argument genome to be of class data.frame. This avoids downloading the refflat file at each call. - Fixed bug in enrichedRegions which caused it to crash when sample1 had more chromosomes than sample2 - Fixed bug in filterDuplReads which caused it to crash when no reads needed to be filtered - Fixed bug in regionsCoverage which caused it to crash for empty chromosomes - Removed functions genePlot, rangesPlot (moved to package casper) - Added regionsCoverage, gridCoverage, stdGrid functions - Added chipSeqQC function - enrichedRegions was trying to load multicores library instead of multicore. - Added ssdCoverage function - Added enrichedRegions method for RangedDataList objects. This allows comparing proportion of reads in islands across >2 samples. - Added islandCounts function - Improved memory usage and speed in filterDuplReads, which no longer relies on countOverlaps from package IRanges - Changed call to parallel to multicore::parallel to avoid confusion with other packages VERSION 0.2.5 - Added arguments maxRepeats and ppOverAmp to filterDuplReads to allow it to eliminate reads appearing more than k times (before only k=1 was allowed), or with a large posterior probability of being an artifact - Added ppEnrichedCounts function - Modified enrichedRegions to make it compatible with new IRanges version - Added cmds function to perform MultiDimensional Scaling on sequencing data - Exact P-value calculation in enrichedRegions is now based on Fisher's exact test, instead of simulations, to increase speed. Changed parameters simulate.p.value and B for exact. - Chi-square P-value calculation in enrichedRegions is now substantially faster (reduced to about 25% of previous implementation) - enrichedRegions now selects regions with pval <= pvalFilter, instead of pval < pvalFilter. This allows to obtain all regions by setting pvalFilter=1. - enrichedRegions now returns reads per kilobase per million (RPKM), as well as raw counts - Corrected enrichedRegions fold change calculation. Now it is (counts1/nsample1)/(counts2/nsample2) instead of (counts1/(nsample1-counts1))/(counts2/(nsample2-counts2)) - Reduced substantially the amount of memory required by rangesPlot VERSION 0.2.3 - Fixed issue in enrichedRegions which caused it to return an object with too many spaces - rangesPlot now produces a warning when no ranges are found in the specified region, but it still produces a plot - Added genePlot method for signature 'RangedData' - Fixed issue in genePlot which caused it to ignore the ... argument. - Fixed issue in genePlot which sometimes caused it to plot additional splicing variants - Fixed issue in 'filterDuplReads' which caused the returned object to have a too large number of spaces - Added methods for signature(regions='RangedData', sample1='RangedData', sample2='missing') and - Fixed bug in enrichedRegions when no hits were found. Also changed argument 'parallelComp' for 'mc.cores', so facilitate using mclapply instead of 'parallel' calls - Added function tabDuplReads - Fixed bug in filterDuplReads for RangedData signature which caused it to filter two reads from different chromosomes which had the same start and end positions - Added extendRanges and filterDuplReads methods for RangedDataList objects. Added mc.cores argument to allow for parallel computing. - Added export2SGR method for RangedDataList objects. - export2SGR now uses 'coverage' function instead of 'pileup', as the latter has been deprecated - Added export2aligned method for RangedDataList objects. Added argument dir. - alignPeaks now returns an object of the same class as its input, instead of always returning IRangesList objects. - Added alignPeaks method for RangedDataList objects, and parameter mc.cores to allow parallel computation. - Major change in readAlignedBatch. Sequences from all files are now returned into a single RangedDataList object. Files are not directly saved to disk as in previous versions. - Fixed bug in rangesPlot which caused it to plot reads longer than maxFragLength twice - Fixed bug in genePlot which caused an error for genes with a single exon - Added PeakLocation function. It produces the same plots as stdPeakLocation except that it doesn't standardize the coordinates. - Added p-value adjustment in enrichedRegions. Defined methods for the case when sample2 is missing. - readAlignedBatch now checks for the existence of the files, and returns a warning if they don't. - Added listOverlap function, which allows to assess the overlap between lists e.g. two ChIP-seq experiments, or ChIP-Seq vs microarray results. - Corrected bug in alignPeaks which could cause sequences to start at negative indexes. VERSION 0.2.1 - ReadAlignedBatch now returns 'RangedData' objects. This required substantial structural changes, as new methods had to be defined and many existing ones needed to be adjusted. - Added methods 'filterDuplReads' and 'mergeRegions' for signature 'RangedData' - Changed stdPeakLocation to a generic function and defined methods for signatures data.frame and RangedData. - Added closestGene method for signature 'RangedData'. - Added enrichedPeaks function to detect peaks in sequencing experiments. - Added enrichedRegions function to detect significantly enriched regions in sequencing experiments. - closestGene modified: After bioC upgrade parameter multiple stoped working due to findOverlaps function that deprecated Overlaps function (which was used in previous versions). - closestGene: Some rbinds have been removed and do.call is used insted to use less memory. - closestGene: Better control of downloaded refflat files has been added. - closestGene: new parameter added (promoterInOther). If it's TRUE peaks that are in a promoter region and inside other gene will be assigned to both genes. Only works when multiple=TRUE. VERSION 0.2.0 - Fixed bug in alignPeaks which caused it not to work when there were some chromosomes with no sequences. Also, now alignPeaks reports the estimated shift. - readAlignedBatch deletes memory consuming objects after they have been saved to disk. This lowers memory requirements. - readAlignedBatch forces the chromosome names to start by 'chr', to ensure compatibility with external softwares. - Added vignette - Added Rd file for alignPeaks - Added export2aligned, export2SGR and rangesPlot methods for RangedData and RangedDataList objects. - export2SGR now longer requires to specify chromosome length for IRangesList and RangedDataList objects. The range of positions where reads have been observed is used instead. VERSION 0.1.2 - Added vignette - Exported rangesPlot in NAMESPACE. - Added function stdPeakLocation to plot peak location in ChIP-Seq and ChIP-chip experiments - Corrected mismatching arguments in generic definition for genePlot. Also did minor changes to improve label appearance. - closestGene did not work after upgrading to R 2.10. VERSION 0.1.1 - closestGene now uses findOverlaps function instead of the deprecated 'overlap'. - Converted rangesPlot function to S4. Added methods for character, IRanges and IRangesList. - Converted genePlot function to S4. Added methods for IRanges, IRangesList and CompressedIRangesList objects (additionally to the already existing one for character). - Added alignPeaks function to align the '+' and '-' reads in ChipSeq experiments - Now readAlignedBatch returns a RangeDataList object for single end experiments instead of an IRangesList. This allows to store the strand information, which could not be done with the IRangesList object. - Added export2SAM function - Added export2aligned method for 'IRanges' objects - Added 'paired' option in readAlignedBatch - Changed default type='Bowtie' in readAlignedBatch - Added genePlot function - closestGene had an error when looking if a region falls in a promoter. - Distance to gene added to clsoestGene function. VERSION 0.1.0 - Added readAlignedBatch function VERSION 0.0.1 - extendRanges now outputs an IRangesList object instead of a simple list - Added export2aligned method. - Added export2SGR method for IRangesList objects - Added closestGene function. - Changed "c" for "rbind" in "enrichedRegions.r" line 92 (needed for IRanges version 1.6.0) - Added require(multicore) to "enrichedRegions.r" and changed mclapply to lapply for non-parallel computation