Changelog

Source: NEWS.md

ggnomics 0.1.2

Major changes

• Started with a clean slate, removed previous ggnomics content.
• Previous functionality unrelated to genomics now lives in the ggh4x GitHub.

Plotting

• Added fortify methods for S4Vectors::DataFrame and S4Vectors::Vector classes, preserving S4 classes where possible.
• Added ggplot() method for S4Vectors::Vector class, which automatically adds coord_S4().

Geoms

• Added geom_range(), which is effectively geom_rect(), but can be parameterised by IRanges::Ranges-class for the x/y aesthetic instead of the xmin/xmax and ymin/ymax parameterisation.
• Added geom_genemodel() to display exon/intron structure of genes. Controlling the thickness of CDSs vs UTRs can be done using scale_exontype_discrete().

Stats

• Added stat_coverage() to calculate how often a space is covered by a set of ranges. Treats IRanges::Ranges-class as closed-interval integers and doubles as open-interval.

Positions

• Added position_disjoint_ranges() to segregate overlapping ranges. Works similar to IRanges::disjointBins(), but doesn’t require integer ranges.

Scales

• Added tree of ‘sister’ classes for ggplot’s discrete and continuous scales. It is the task of a scale to take S4 classes as input and wrap them inside an S3 representation and expose the correct training (range updating).
• Added scale_(x/y)_S4_continuous() to handle continuous S4-class positions.
• Added scale_(x/y)_S4_discrete() to handle discrete S4-class positions.
• Added scale_(colour/fill)_S4_continuous() and scale_(colour/fill)_S4_discrete() to handle S4-class values.
• Added scale_(x/y)_genomic(), which is almost identical to the scale_(x/y)_S4_continuous() scale except for labels, breaks and guide defaults.
• Added guide_genomic_axis() to display sequence name - position relation more clearly.
• Dispatch rules for scale types; for S4Vectors::Factor and S4Vectors::Rle, the levels/value slot is inspected to decide. Genomic classes in the GenomicRanges package go to *_genomic scales. Anything else is currently defaulting to the *_S4_continuous scales.
• Added guide_ideogram_axis() to annotate positions with ideograms. Not implemented for genomic scales yet!

Coordinates

• Added coord_S4() to control class specific final transformations. This coordinate system is added automatically when ggplot() is called on a S4Vectors::Vector class. Should be added manually when S4 Vector data is added in, for example, layers. This allows the plot to ‘know’ that e.g. an IRanges::IRanges object underneath an xmin aesthetic should give the start, underneath xmax should give the end and under a plain x should give the middle.

Internals

• Added custom ViewScale ggprotos that can give minor labels and handle discontinuous limits (for e.g. GenomicRanges::GRanges).
• Added forceflat generic to map discontinuous classes to linear [0,1] interval. For genomic classes, mimicks GenomicRanges::absoluteRanges(); otherwise scales::rescale().
• Added range related generics for getting the range of a class, expanding a range and determining empty ranges.
• Added plot arithmetic for WoodenHorse, so that Ranges obey e.g. nudging rules.
• Added generic out-of-bounds variants for censor(), squish(), squish_infinite() and discard() from the scales package.
• Added major/minor breaks S4BreaksMajor()/S4BreaksMinor() and labels S4LabelFormat() generics. They default to the scales package behaviour, but differ for genomic classes (e.g. GRanges).
• Added S4TransIdentity() identity transformation that uses the major/minor breaks and label generics.
• S3 wrapper-representation of S4 Vector class, so that they are compatible with data.frame operations. The S3 class is called WoodenHorse as they smuggle S4 Vectors into an S3 framework. Constructor is GreekSoldier(), reversor is Nightfall() and inspector of the underlying class in HelenOfTroy(). There are two WoodenHorse variants; BeechHorse which carries the S4 Vector as a parallel attribute, and OakHorse which is an index into the S4 Vector attribute.
• Added a NEWS.md file to track changes to the package.