The source tarball is now under 100 MB (94.8 MB), down from over
100 MB. The following files were removed. All of them are archived, with the full
Salmon output for all 24 samples, on Zenodo:
https://doi.org/10.5281/zenodo.22982581. The with-inf-reps branch
on GitHub is a frozen snapshot of the package that still includes them.
inst/extdata/quants/*/aux_info/bootstrap/, ~393 MB).
tximeta() / tximport() on the files in inst/extdata/quants no
longer imports inferential replicates.aux_info/ambig_info.tsv.gz for each sample.inst/extdata/gencode.v29.annotation.gtf.gz.inst/extdata/gencode.v29.annotation.slim.gtf.gz: a slim Gencode v29
GTF with only gene and transcript lines and 6 attributes (5 MB instead
of 38 MB). It has no exon lines, so tximeta::addExons() will not
work with it. It has a different filename from the full GTF on purpose:
tximeta caches TxDbs by GTF filename, so this avoids clashing with a
TxDb built from the full Gencode v29 GTF. Code that called
makeLinkedTxome(gtf=...) with the full GTF path should switch to the
slim filename. Built by inst/scripts/makeSlimGTF.sh.data(macro_txp_se): a transcript-level SummarizedExperiment with
20 inferential replicates, for all transcripts of 2000 random expressed
genes plus GBP1-GBP7. The counts and infRep assays are rounded to
1 decimal. This is now used by the fishpond swish vignette.data(macro_tx2gene): a transcript-to-gene table for all Gencode v29
transcripts, for use with
tximeta::summarizeToGene(se, skipRanges=TRUE, tx2gene=macro_tx2gene).inst/scripts/makeSumExp.R. The script
needs a macrophage install that still has the inferential replicates
(<= 1.29.0, or the with-inf-reps branch).data(gse) object and the quantification files
(quant.sf.gz, etc.) in inst/extdata/quants are unchanged.