Changes in version 0.99.2                        

    o   The conversion functions are renamed in camelCase, following the
	naming of `SeqArray::seqVCF2GDS()`: `seqToGDS_FAVOR()` =>
	`seqFAVOR2GDS()`, `seqToGDS_FAVOR_tar()` => `seqFAVORTar2GDS()`,
	`seqToGDS_gnomAD()` => `seqGnomAD2GDS()`, `seqToGDS_VEP()` =>
	`seqVEP2GDS()`, `seqToGDS_SnpEff()` => `seqSnpEff2GDS()` and
	`seqToGDS_ANNOVAR()` => `seqANNOVAR2GDS()`

    o   New help page `?GDSAnnotator-extdata` describing the example data;
	'inst/script' is renamed to 'inst/scripts'

    o   Internal code reorganized into smaller functions, without changing the
	outputs; the benchmark script is removed from 'inst'

                       Changes in version 0.99.1                        

    o   New functions `seqAnnotStat()`, `seqAnnotReport()` and
	`seqAnnotGeneTable()` for generating a VEP/SnpEff-like summary report
	from GDS files: `seqAnnotStat()` collects all summary statistics in a
	single pass over the file (the counters are additive, so multiple GDS
	files, e.g., one per chromosome, can be summarized together), and
	`seqAnnotReport()` renders them as a report in HTML (plain HTML tables
	and inline SVG charts, no additional package is required), Markdown
	(pipe tables) or R Markdown (the charts are drawn at knit time) format;
	'parallel' splits the variants between multiple processes, and the
	per-process counters are merged afterwards

    o   New function `seqToGDS_ANNOVAR()` for importing ANNOVAR-annotated VCF
	files (from table_annovar.pl -vcfinput) into SeqArray GDS format

    o   new 'bsize' in `seqAnnotate()` and `seqAnnotateGDS()` to split the
	variants into blocks when 'add_to_gds' is TRUE or a file name: the
	annotation data are loaded and written block by block, so the memory
	usage no longer depends on the total number of variants (requiring the
	block writing in `SeqArray::seqAddValue()`, SeqArray_v1.53.2)

    o   `seqAnnotate()` accepts the `VCF` objects of the package
	VariantAnnotation (CollapsedVCF and ExpandedVCF), annotating one row
	per alternative allele; `seqAnnotateVCF()` does the same after reading
	the file (VariantAnnotation is a suggested package, loaded on demand)

    o   Bug fix: `seqAnnotate()` failed when 'varnm' was missing for a
	data.frame, DataFrame or IRanges input, instead of using all the
	annotations of the first annotation file

    o   Bug fix: `seqToGDS_FAVOR()` printed progress messages with
	'verbose=FALSE'

    o   The categories of each counter in `seqAnnotStat()` are stored as plain
	named numeric vectors in a fixed order, so that the output is identical
	regardless of 'parallel' and 'bsize'

    o   Explicit exports in NAMESPACE; the examples write to `tempfile()`;
	the provenance of the example data in 'inst/extdata' is documented in
	'inst/script'; new package help page `?GDSAnnotator`

                       Changes in version 0.99.0                        

    o   Initial package