Version 0.4 • AcidGenomes

AcidGenomes 0.4.0 (2022-05-04)

Major changes:

S4 classes containing identifier mappins now are set to contain DFrame instead of DataFrame, as the previous definition approach no longer works with Bioconductor 3.15. This applies to Ensembl2Entrez, Entrez2Ensembl, EntrezGeneInfo, Gene2Symbol, HGNC, HGNC2Ensembl, Protein2Gene, and Tx2Gene classes.
downloadGencodeGenome: Added support for Entrez and RefSeq identifiers, which are now defined in mcols of GRanges objects.
Reworked internal code of rtracklayer runner functions. In particular, improved identifier handling for RefSeq GFF and GTF runners, adding support for extraction of Entrez gene identifiers.

Minor changes:

Reformatted code using styler conventions.
Removed internal dependencies on stringr, in favor of base R when possible and stringi package otherwise.
Hardened import of external files using readr package, where applicable. Applies primarily to Entrez, HGNC, and MGI file import.
makeTx2GeneFromFASTA: Added ignoreVersions argument, which is now enabled by default, to match other Tx2Gene functions.
Updated unit tests to cover most recent versions of genome assembly files.

Minor changes:

Minor changes:

Ensembl2Entrez: Simplified format "1:1" handling, based on new code approach used in AcidGSEA package. This now keeps track of original rownames, which is necessary for bcbioRNASeq clusterProfiler R Markdown template.

Minor chagnes:

makeGRangesFromEnsembl: Hardened internal code to suppress spurious warnings from rtracklayer due to masking of download.file function. See issue #71 for details.
Improved temporary directory creation and deletion using tempdir2 and unlink2 internally, which improves support for continuous integration (CI) checks on Windows.

Minor changes:

Compatibility update to provide support for breaking changes introduced with pipette 0.10.0 release.
EntrezGeneInfo: Added column name checks. Renamed xTaxId to taxonomyId.
Migrated all importFrom calls to imports.R file.
Removed BiocParallel bplapply usage.
export: Updated methods to match new conventions defined in pipette.
Improved support for quick detection of NCBI taxonomic group for commonly used model organisms. Previously human and mouse were supported. Added support for zebrafish, rat, worm, fruitfly, and yeast.

Minor changes:

downloadUCSCGenome: Added a manual override for defaulting to "hg38" for Homo sapiens, which has switched over to experimental "hs1" T2T genome build.
mapGencodeToEnsembl: Added support for mapping Mus musculus releases.
Internal Seqinfo generators now default to pulling annotations from Ensembl for GENCODE reference, rather than UCSC. Note that UCSC Seqinfo function is currently broken for hg38, but will be fixed in pending GenomeInfoDb v1.34.8.
Updated unit tests to check against latest genome build releases.

Minor changes:

Migrated requireNamespaces import from AcidBase to goalie.
Updated dependency versions.
Now allowing Ensembl Seqinfo fetch step to fail, to avoid unit test issues with Mus musulus Ensembl 90 GTF file (for bcbioRNASeq).

New functions:

gencodeReleaseHistory: This function scrapes the GENCODE website to return the full release history for either human or mouse genomes.

Minor changes:

currentEnsemblVersion: Fix for breaking change on Ensembl FTP server. The file this function parses has been renamed from current_README to simply README.
mapGencodeToEnsembl: Now using gencodeReleaseHistory internally to dynamically fetch metadata directly from the GENCODE website, instead of relying on an internal CSV mapping file. This helps avoid having to update the package every time a new Ensembl/GENCODE release comes out.

Minor changes:

makeTx2GeneFromGFF, Tx2Gene: Ensure that rownames are defined for RefSeq genome annotations, which are constructed from GenomicRangesList method.