Work with Bioconductor package dependencies

Bioconductor is built using an extensive set of core capabilities and data structures. This leads to package developers depending on other packages for interoperability and functionality. This function extracts package dependency information from biocPkgList() and returns a tidy data.frame that can be used for analysis and to build graph structures of package dependencies.

buildPkgDependencyDataFrame(dependencies = c("strong", "most", "all"), ...)

Arguments

dependencies

character() a vector listing the types of dependencies, a subset of c("Depends", "Imports", "LinkingTo", "Suggests", "Enhances"). Character string "all" is shorthand for that vector, character string "most" for the same vector without "Enhances", character string "strong" (default) for the first three elements of that vector.

...

parameters passed along to biocPkgList()

Value

A data.frame (also a tbl_df) of S3 class "biocDepDF" including columns "Package", "dependency", and "edgetype".

Note

This function requires network access.

See also

See buildPkgDependencyIgraph(), biocPkgList().

Examples

# performs a network call, so must be online.
library(BiocPkgTools)
depdf <- buildPkgDependencyDataFrame()
#> 'getOption("repos")' replaces Bioconductor standard repositories, see
#> 'help("repositories", package = "BiocManager")' for details.
#> Replacement repositories:
#>     CRAN: https://cran.rstudio.com
head(depdf)
#>   Package  dependency edgetype
#> 1      a4      a4Base  Depends
#> 2      a4   a4Preproc  Depends
#> 3      a4   a4Classif  Depends
#> 4      a4      a4Core  Depends
#> 5      a4 a4Reporting  Depends
#> 6  a4Base   a4Preproc  Depends
library(dplyr)
# filter to include only "Imports" type
# dependencies
imports_only <- depdf |> filter(edgetype=='Imports')

# top ten most imported packages
imports_only |> select(dependency) |>
  group_by(dependency) |> tally() |>
  arrange(desc(n))
#> # A tibble: 1,877 × 2
#>    dependency               n
#>    <chr>                <int>
#>  1 stats                 1328
#>  2 methods               1260
#>  3 utils                 1083
#>  4 ggplot2                743
#>  5 S4Vectors              672
#>  6 grDevices              596
#>  7 graphics               595
#>  8 dplyr                  544
#>  9 SummarizedExperiment   498
#> 10 IRanges                451
#> # ℹ 1,867 more rows

# The Bioconductor packages with the
# largest number of imports
largest_importers <- imports_only |>
  select(Package) |>
  group_by(Package) |> tally() |>
  arrange(desc(n))

# not sure what these packages do. Join
# to their descriptions
biocPkgList() |> select(Package, Description) |>
  left_join(largest_importers) |> arrange(desc(n)) |>
  head()
#> 'getOption("repos")' replaces Bioconductor standard repositories, see
#> 'help("repositories", package = "BiocManager")' for details.
#> Replacement repositories:
#>     CRAN: https://cran.rstudio.com
#> Joining with `by = join_by(Package)`
#> # A tibble: 6 × 3
#>   Package      Description                                                     n
#>   <chr>        <chr>                                                       <int>
#> 1 singleCellTK "The Single Cell Toolkit (SCTK) in the singleCellTK packag…    85
#> 2 ChromSCape   "ChromSCape - Chromatin landscape profiling for Single\nCe…    57
#> 3 signeR       "The signeR package provides an empirical Bayesian approac…    54
#> 4 metaseqR2    "Provides an interface to several normalization and\nstati…    52
#> 5 FLAMES       "Semi-supervised isoform detection and annotation from bot…    51
#> 6 scRNAseqApp  "The scRNAseqApp is a Shiny app package designed for\ninte…    50