A set of tools to create a UK Biobank < http://www.ukbiobank.ac.uk/> dataset from a UKB fileset (.tab, .r, .html), visualize primary demographic data for a sample subset, query ICD diagnoses, retrieve genetic metadata, read and write standard file formats for genetic analyses.
If you use ukbtools in any published work, please cite:
Hanscombe K.B., Coleman J.R.I., Traylor M., and Lewis C.M. ukbtools: An R package to manage and query UK Biobank data. bioRxiv 158113; doi: https://doi.org/10.1101/158113
After downloading and decrypting your UK Biobank (UKB) data with the supplied UKB programs, you have multiple files that need to be brought together to give you a dataset to explore. The data file has column names that are edited field-codes from the UKB data showcase. ukbtools makes it easy to collapse the multiple UKB files into a single dataset for analysis, in the process giving meaningful names to the variables. The package also includes functionality to retrieve ICD diagnoses, explore a sample subset in the context of the UKB sample, and collect genetic metadata.
# Install from CRANinstall.packages("ukbtools")# Install latest development versiondevtools::install_github("kenhanscombe/ukbtools", dependencies = TRUE)
Note: This package is in beta - it is feature complete but may contain unknown bugs. If anything does not work, first re-install the package
devtools::install_github("kenhanscombe/ukbtools", dependencies = TRUE, force = TRUE) to get the latest development version. If it is still not working, let me know and I'll fix it.
Download§ then decrypt your data and create a "UKB fileset" (.tab, .r, .html):
ukb_unpack ukbxxxx.enc keyukb_conv ukbxxxx.enc_ukb rukb_conv ukbxxxx.enc_ukb docs
ukb_unpack decrypts your downloaded
ukbxxxx.enc file, outputting a
ukb_conv with the
r flag converts the decrypted data to a tab-delimited file
ukbxxxx.tab and an R script
ukbxxxx.r that reads the tab file. The
docs flag creates an html file containing a field-code-to-description table (among others).
§ Full details of the data download and decrypt process are given in the Using UK Biobank Data documentation.
ukb_df() takes two arguments, the stem of your fileset and the path, and returns a dataframe with usable column names. This will take a few minutes. The rate-limiting step is reading and parsing the code in the UKB-generated .r file - not
ukb_df per se.
library(ukbtools)my_ukb_data <- ukb_df("ukbxxxx")
You can also specify the path to your fileset if it is not in the current directory. For example, if your fileset is in a subdirectory of the working directory called data
my_ukb_data <- ukb_df("ukbxxxx", path = "/full/path/to/my/data")
Note: You can move the three files in your fileset after creating them with
ukb_conv, but they should be kept together.
ukb_df() automatically updates the read call in the R source file to point to the correct directory (the current directory by default, or a directory specified by
All tools are described on the ukbtools webpage and in the package vignette "Explore UK Biobank Data"
vignette("explore-ukb-data", package = "ukbtools")
For a list of all functions
help(package = "ukbtools")
ukb_df: Replaced readr::read_tsv with data.table::fread for faster read. Also includes an
n_threadsargument passed to
data.table::fread, which may make read faster. Column names now include field code to ensure names are unique (UK Biobank sometimes use the same description for more than one variable)
ukb_defunctexplains why these have become defunct and where to get UK Biobank genetic (meta)data.
ukb_gen_sqc_namessupplies column names for the separately downloaded sample QC file;
ukb_gen_rel_countdoes the same as before (a count of levels of relatedness or a plot) but with separately downloaded relatedness data;
ukb_gen_related_with_datareturns subset of relatedness data in which both IDs have data on a phenotype of interest;
ukb_gen_samples_to_removereturns a list of individuals to exclude in order to remove relatedness (one possible solution to a maximal subset problem).
ukb_icd_freq_by: corrected order by levels of
reference.var in the optional plot. (order in the default dataframe returned was correct.)
ukb_df: corrected tab file path update in r source file. Specifically, made regular expression more specific (1 case reported of regular expression matching word elsewhere in the source file.). Also, replaced utils::read.delim with readr::read_tsv for faster read, with progress bar.
ukb_icd_freq_by returns frequency for one or more ICD diagnoses by levels of a reference variable and includes an optional plot
ukb_df_full_join (a thin wrapper around
dplyr::full_join) recursively called on a list of UKB datasets
ukb_df_duplicated_names to identify duplicated names within a dataset. The variable prefix (constructed from its description), index, and array should make the column name unique. However, typos in UKB documentation that give two variables the do not increment index/array have been observed. You will want to identify these and update them appropriately. We expect the occurrence of such duplicates will be rare.
ukb_icd_diagnosis now takes one or more individual ids and returns a dataframe with a potential message noting ids with no diagnoses
ukb_icd_keyword accepts a character vector of one or more "keywords" and returns all ICD descriptions including any of the keywords