CHANGES IN VERSION 1.4.0
-------------------------

NEW FEATURES

    o BiomartGeneRegionTracks will now make use of the available CDS
      information in Ensembl.

    o The constructors to the AnnotationTrack, GeneRegionTrack,
      DataTrack and SequenceTrack classes now accept a character
      scalar that points to a file on the file system. A number of
      default parser functions have been implemented to read the
      standard file types. Alternatively, a user-defined import
      function can be provided. This feature also supports streaming
      from indexed file types like BAM or bigWig, in which case the
      data is fetched dynamically upon each plotting operation.

    o The mart object in BiomartGeneRegionTrack objects is now cached
      in order to speed up subsequent queries to the same mart.

    o When plotting DataTracks with type 'gradient' or 'heatmap', a
      color scale is plotted next to the regular y-axis to indicate
      the mapping of numeric values in the false color range. Thanks
      to Mark Heron for his code contribution.

    o Sample names can now be shown in heatmap-type plots by setting
      the 'showSampleNames' display parameter.

SIGNIFICANT USER-VISIBLE CHANGES

    o Complete refactoring of the automatic font size adjustments to
      provide more reasonable defaults.

    o Tick labels on the genomic axis are now show in between tick
      marks when zoomed in to single nucleotide level.

BUG FIXES

    o Fixed a bug in IdeogramTracks where all bands in the rounded
      caps at the edges of the Ideogram were missing.

    o The way genomic ranges are plotted is now according to the Lego
      block model suggested by Herve. This is only relevant when zooming
      in to the level of single nucleotides.

    o Tick labels on the genome axis show only significant digits now.

    o Sample ordering in heatmap plots is now correct.

    o Numerous little fixes.

CHANGES IN VERSION 1.2.0
-------------------------

NEW FEATURES

    o A SequenceTrack class has been added to draw genomic sequence
      information on a Gviz plot. Possible inputs for the track are
      DNAStringSet objects or directly from BSgenome packages.

    o GeneRegionTracks can now deal with coding and non-coding regions
      by means of the feature property in combination with the
      thinBoxFeature display parameter.

    o StackedTracks now have a new display parameter 'reverseStacking'
      which reverts the horizontal ordering of stacked items. If set
      to TRUE, the lowest items are moved to the top of the stack, and
      vice versa.

SIGNIFICANT USER-VISIBLE CHANGES

    o Updated the show methods for most tracks to give more meaningful
      and more compact information about the track's
      content. Availablability of data on other chromosomes than the
      currently active one should now be indicated.

    o IdeogramTracks can now be constructed from a cytoband table via
      the new bands argument in the constructor.

    o AnnotationTrack objects now by default draw connecting lines in
      a light gray color. This feature can be controlled via the
      col.line display parameter.

    o Sliding window summarization can now deal with NA values.

    o Exporting drawGD from the name space now to allow for
      sub-classing of GdObjects in other packages.

    o When building GeneRegionTracks from TranscriptDb objects the
      information about UTRs and coding regions is now retained.

   
BUG FIXES

    o When zooming into the emty space between two grouped features,
      the connecting line will now be plotted for all classes inheriting
      from AnnotationTrack.

    o An error in calculating ylims when drawing AlignedReadTracks has
      been fixed.

    o Numerous other little fixes that mainly aim at improving
      performance.