extract_biomarker_qc_flags {ukbnmr}

Extract NMR biomarker QC flags from a data.frame of UK Biobank fields

Description

Given an input data.frame loaded from a dataset extracted by ukbconv extracts the UK Biobank fields corresponding to the quality control (QC) flags for the NMR metaolomics biomarkers giving them short variable names as listed in the nmr_info information data sheet available in this package. QC Flags are separated by "; " in each column where there are multiple QC Flags for a single measurement.

Usage

extract_biomarker_qc_flags(x)

Arguments

x

data.frame with column names "eid" followed by extracted fields e.g. "23774-0.0", "23774-1.0", ..., "23767-1.0".

Details

ukbconv have one row per UKB biobank participant whose project specific sample identifier is given in the first column named "eid". Columns following this have the format "<field_id>-<visit_index>.<repeat_index>", where here <field_id> corresponds to a biomarker of interest, e.g. 23774 for the QC Flags for 3-Hydroxybutyrate, and <visit_index> corresponds to the assessment time point, e.g. 0 for baseline assessment, 1 for first repeat visit.

The <repeat_index> field is currently used by UK Biobank to index cases where there are multiple QC Flags for a biomarker for a single participant and visit_index. QC Flags are separated by "; " in each column where there are multiple QC Flags for a single measurement.

In the returned data.frame there is single column for each biomarker, with an additional column for the visit index. Rows are uniquely identifiable by the combination of entries in columns "eid" and "visit_index".

If your UK Biobank project only has access to a subset of biomarkers, then this function will only return the subset of ratios that can be computed from the biomarker data provided.

Value

a data.frame or data.table with column names "eid", and "visit_index" followed by columns for each biomarker e.g. "bOHbutyrate", ..., "Valine".

Examples

ukb_data <- ukbnmr::test_data # Toy example dataset for testing package
biomarker_qc_flags <- extract_biomarker_qc_flags(ukb_data)