vforget / ukb

UK Biobank Tools

UK Biobank Tools

Scripts to parse and analyze UK Biobank data.

melt_tab.awk

Awk script to melt tab-delimited UK Biobank dataset, similar the melt function of the reshape2 R package.

The motivation for this script is to faciliate the convertion of large tab-delimited UK Biobank dataset to long format, or to melt the dataset. For a small dataset, please consider using the melt function of the reshape2 R package as it offers greater functionality.

Usage

awk -f melt_tab.awk ukb####.tab > ukb####.melt

Output

Missing values (NA) are discarded.

Columns in tab-delimited output are:

1. eid: Participant id.
2. df: Data-field id.
3. instnum: Instance number for given data-field (0-based).
4. itemnum: Item number for given instance (0-based).
5. value: Data-field value for given participant, instance number and item number.

fetch_tab.awk

Gawk script to subset tab-delimited UK Bionank dataset by list of user-supplied data-fields.

Usage

gawk -f fetch_tab.awk data-fields.txt ukb####.tab | gzip -c > ukb####_fetch.tab.gz

File data-fields.txt contains data-field identifiers, one per line. Example:

21022
52
34
31

Output

Output is a subset of input UK Biobank dataset by the supplied data-fields. The format of the file remains tab-delimited, and data-fields retain original order. For each data-field, all instance and items are output. All participants are output.