Changes in version 1.41.1                        

Reduced package size

The package's example data has been cut from ~593 MB on disk (~388 MB
compressed) to ~171 MB on disk (~96 MB compressed). The built tarball is
now ~96 MiB.

The full data as it was before this change (commit 02887f7) is kept on
the with-inf-reps branch and tag at
https://github.com/mikelove/tximportData and archived on Zenodo:
https://doi.org/10.5281/zenodo.22982575

Removed

  - salmon_gibbs/: Salmon 0.8.1 output with Gibbs samples (Ensembl v87,
    6 samples)
  - kallisto_boot/: kallisto 0.43.0 output with bootstraps (Ensembl v87,
    6 samples)
  - tx2gene.ensembl.v87.csv
  - gencode/gencode.v48.annotation.gtf.gz. tximeta can now link
    transcripts to genes with a DataFrame, so it no longer needs a GTF.
  - alevin/{mouse1_unst_50,mouse1_LPS2_50}/alevin/bfh.txt and
    raw_cb_frequency.txt
  - kallisto/*/abundance.h5 (no bootstraps). abundance.tsv.gz is kept.
  - GEUVADIS samples ERR188297, ERR188329, ERR188288 and ERR188356 from
    kallisto/, rsem/, salmon/ and sailfish/. samples.txt and
    samples_extended.txt now have 2 rows: ERR188088 and ERR188021.

Kept

  - salmon_ec/*/aux_info/eq_classes.txt (used by fishpond)
  - alevin/mouse1_unst_50_boot/ (used by eds)
  - cufflinks/, still with all 6 samples

Changes needed in downstream code

  - Code that reads samples.txt now gets 2 samples instead of 6. For
    example, replace paste0("sample", 1:6) with paste0("sample",
    seq_along(files)), and replace rep(c("A","B"), each=3) with
    factor(c("A","B")).
  - For example data with inferential replicates (Gibbs samples or
    bootstraps), use the Zenodo record above or another data package
    such as macrophage.