Changes in version 1.9.9:

    o   Added extractPatch() function to count bin pairs in a specified area
	 of the interaction space.

    o   Modified connectCounts() to eliminate warnings upon stranded
	 entries, unknown chromosomes. All entries of input regions are now
	 retained, though not necessarily in the input order. Also switched
	 original metadata to NA when second.regions is an integer.

    o   Modified preparePairs() to be more generous when considering
	 inward-facing reads if they overlap past each other.

    o   Fixed bug in savePairs() involving failure to swap other information
	 when enforcing index ordering.

    o   Added mergeCMs() function to allow entry into the pipeline from
	 ContactMatrix objects.

    o   Moved pre-processing scripts out of the package to the repository
	 for the user's guide.

    o   Updated presplit_map.py to use new samtools API for sorting.

    o   Updated user's guide.

Changes in version 1.8.0:

    o   Streamlined filterDirect(), filterTrended(), and added tests for
	 them. Also allowed specification of which assay to use for the data
	 and reference objects.

    o   enrichedPairs() and neighborCounts() now return counts for
	 neighbourhood regions, not just the enrichment values.

    o   filterPeaks() will compute (and optionally return) enrichment values
	 from neighbourhood counts.

    o   normalizeCNV() and correctedContact() allow specification of which
	 assay matrix to use from the SE objects.

    o   Refactored a great deal of the C++ code for improved clarity.

    o   Overhauled handling of DNase Hi-C data, so that pseudo-fragments are
	 no longer necessary. Most functions now automatically recognise
	 DNase-C data from an empty GRanges in param$fragments.  Deprecated
	 segmentGenome() and prepPseudoPairs(), added the emptyGenome()
	 function.

    o   Updated user's guide.

Changes in version 1.6.0:

    o   Relaxed checks in preparePairs(), prepPseudoPairs() when more
	 chromosomes are supplied than needed.

    o   Relaxed checks in connectCounts() when extra chromosomes are in the
	 input ranges.

    o   Fixed an issue with preparePairs() when too many file handles are
	 open.

    o   Fixed clash between BiocGenerics and Matrix which().

    o   Modified domainDirections() to return a RangedSummarizedExperiment.

    o   Removed defunct DIList class and methods.

    o   Switched from seqlevels() to seqlevelsInUse() for fragment
	 intervals.

    o   Updated user's guide, documentation.

Changes in version 1.4.0:

    o   Deprecated DIList objects and methods in favour of InteractionSet
	 objects.

    o   marginCounts() now returns a RangedSummarizedExperiment for all
	 bins.

    o   Added the max.height argument to the rotPlaid() and rotDI()
	 functions.

    o   Added the diClusters() function for post-hoc cluster-level FDR
	 control.

    o   Added the annotatePairs() function for convenient annotation of
	 (clusters of) interactions.

    o   Fixed a bug in plotPlaid() when the interaction space was empty.

    o   Fixed a bug in preparePairs() where unmapped chimeric segments led
	 to the loss of the entire pair.

    o   Updated user's guide, documentation and tests.

Changes in version 1.2.0:

    o   Renamed normalize() to normOffsets().

    o   Added library size specification to DIList methods normOffsets(),
	 asDGEList().

    o   Fixed bugs under pathological settings in plotPlaid(), plotDI(),
	 rotPlaid(), rotDI().

    o   Optimized C++ code for connectCounts(), squareCounts().

    o   Streamlined various R utilities used throughout all functions.

    o   Added iter_map.py to inst/python, for iterative mapping of DNase
	 Hi-C data.

    o   Added the neighborCounts() function, for simultaneous read counting
	 and enrichment calculation.

    o   Added exclude for enrichedPairs(), to provide an exclusion zone in
	 the local neighborhood.

    o   Switched default colour in rotPlaid(), plotPlaid() to black.

    o   Added compartmentalize() function to identify genomic compartments.

    o   Added domainDirections() function to help identify domains.

    o   Modified correctedContact() to allow distance correction and report
	 factorized probabilities directly.

    o   Modified marginCounts() function for proper single-end-like
	 treatment of Hi-C data.

    o   Extended clusterPairs() to merge bin pairs from multiple DILists.

    o   Switched to reporting ranges directly from boxPairs(), added support
	 for minimum bounding box output.

    o   Modified consolidatePairs() to accept index vectors for greater
	 modularity.

    o   Added reference argument for large bin pairs, in filterDirect() and
	 filterTrended().

    o   Added filterDiag() convenience function for filtering of
	 (near-)diagonal bin pairs.

    o   Slight change to preparePairs() diagnostic reports when dedup=FALSE,
	 and for unpaired reads.

    o   Added option for a distance-based threshold to define invalid
	 chimeras in preparePairs().

    o   Updated documentation, tests and user's guide.

    o   Added diffHic paper entry to CITATION.

Changes in version 1.0.0:

    o   New package diffHic, for detection of differential interactions from
	 Hi-C data.