R Client

The R client is a reticulate wrapper around the python client. To get started, first install and configure biobricks:

Installation

$ pip install biobricks
$ biobricks configure

Then install the R client:

> library(remotes)
> install_github("biobricks-ai/biobricks-r",dep=TRUE)
> biobricks::bblib() # check biobricks library path

The R client has a dependency on arrow which is necessary for loading parquet files. This is why the dep=True is set in the install_github command. If you already have arrow installed, the dep argument can be removed.

On the first usage, the Biobricks R package asks to install miniconda, it uses miniconda to install a copy of the python biobricks package. The biobricks-r functions use the reticulate package to call the python biobricks functions.

Loading Bricks

To install and start using a brick use bbinstall and bbload. The clinvar brick distributes parquet files. This means the arrow package is required to load the brick.

> biobricks::bbinstall("clinvar") # install the clinvar brick
> cv <- biobricks::bbload("clinvar")
> cv$variant_summary |> head() |> dplyr::collect()
# A tibble: 6 x 34
#   AlleleID Type     Name  GeneID GeneS~1 HGNC_ID ...
# *    <dbl> <chr>    <chr>  <dbl> <chr>   <chr>   ...
# 1    15041 Indel    NM_0~   9907 AP5Z1   HGNC:2~ ...
# 2    15041 Indel    NM_0~   9907 AP5Z1   HGNC:2~ ...
# ...

The R arrow package and biobricks work really well together, arrow can be used to make analysis of biobricks tables much more efficient, all without pulling data into the R session. Learn more about arrow in R here: https://arrow.apache.org/docs/r/articles/arrow.html