Skip to contents

AcidGenomes 0.5.0 (2023-04-27)

Starting a new release series to denote potential breaking changes with legacy objects saved with entrezId instead of ncbiGeneId.

New functions:

  • Added utility functions to easily map gene symbols to stable gene identifiers at Ensembl, HGNC, and NCBI (Entrez): mapGeneNamesToEnsembl, mapGeneNamesToHGNC, mapGeneNamesToNCBI. These are covered against Homo sapiens and Mus musculus.

Major changes:

  • Classes that extend GRanges (e.g. EnsemblGenes, EnsemblTranscripts, GencodeGenes, GencodeTranscripts) now intentionally fail class checks if entrezId is defined instead of ncbiGeneId in mcols metadata. This makes downstream handoff to GSEA functions in AcidGSEA easier to manage. For legacy objects, use updateObject to resolve this check.
  • makeGRangesFromEnsembl and makeGRangesFromGFF now attempt to fetch additional useful gene metadata, including gene synonyms from the Ensembl FTP server when applicable. This currently applies to gene annotation files from Ensembl and GENCODE. Note that extra metadata is not supported for legacy Homo sapiens GRCh37 genome build.

Minor changes:

  • Renamed EntrezGeneInfo function to NcbiGeneInfo.
  • Removed geneSynonyms from NAMESPACE. Consider using NcbiGeneInfo or HGNC instead for synonym information.
  • Removed HGNC2Ensembl and MGI2Ensembl. Just use HGNC and MGI function return instead.
  • downloadEnsemblGenome now downloads additional useful metadata files.
  • Added updateObject support to update legacy objects that may fail new entrezId class checks.

AcidGenomes 0.5.1 (2023-08-01)

Major changes:

  • downloadGencodeGenome: This genome download function now sanitizes the transcriptome FASTA to only include transcript identifiers in the header without additional information separated by the "|" (pipe) delimeter. This approach is not commonly used in FASTA files, and results in unwanted downstream behavior when quantifying at transcript level using kallisto and minimap2. Note that salmon can currently handle this edge case when setting the --gencode flag during the genome index step. This action is non-destructive and returns a “transcriptome_fixed” FASTA file. We are now symlinking this fixed file by default, but the unmodified original is retained in the transcriptome download folder.

Minor changes:

  • currentEnsemblGenomeBuild: Fixed internal REST API query to Ensembl server. This now requires "content-type=application/json" to be defined in the URL, otherwise the Ensembl server returns text instead of JSON.
  • currentEnsemblVersion: Now parses current_README file on FTP server instead of the top level README. We changed this because the README symlink can break during Ensembl release updates (e.g. 109 to 110, in progress).