UK Biobank NMR

https://github.com/sritchie73/ukbnmr
Admin
UK Biobank NMR is an R package providing utilities for processing UK Biobank Nuclear Magnetic...
Tokens:6,990
Snippets:58
Trust Score:8.2
Update:7 months ago
Show doc for...
Tokens:
Raw
### Install ukbnmr package

Source: https://context7.com/sritchie73/ukbnmr/llms.txt

Instructions for installing the ukbnmr package from CRAN or the development version from GitHub.

```R
# Install from CRAN
install.packages("ukbnmr")

# Install latest development version from GitHub
remotes::install_github("sritchie73/ukbnmr", dependencies = TRUE, build_vignettes = TRUE)
```

--------------------------------

### Install ukbnmr R Package from GitHub

Source: https://github.com/sritchie73/ukbnmr/blob/main/README.md

Installs the most up-to-date version of the ukbnmr R package directly from its GitHub repository. This method ensures access to the latest features and bug fixes. Dependencies are automatically installed, and vignettes are built.

```r
remotes::install_github("sritchie73/ukbnmr", dependencies = TRUE, build_vignettes = TRUE)
```

--------------------------------

### Install ukbnmr R Package from CRAN

Source: https://github.com/sritchie73/ukbnmr/blob/main/README.md

Installs major releases of the ukbnmr R package directly from the Comprehensive R Archive Network (CRAN). This is a stable and recommended method for general use.

```r
install.packages("ukbnmr")
```

--------------------------------

### Get Description for Specific Field in R

Source: https://context7.com/sritchie73/ukbnmr/llms.txt

Filters the sample quality control information to retrieve the description for a specific field, identified by its name. This helps in understanding individual QC parameters.

```r
sample_qc_info[Name == "Spectrometer"]
```

--------------------------------

### View Sample QC Fields in R

Source: https://context7.com/sritchie73/ukbnmr/llms.txt

Prints a table containing sample quality control information, including field names, descriptions, and UKB Field IDs. This is useful for understanding the available QC metrics.

```r
print(sample_qc_info)
```

--------------------------------

### Load and Inspect Test Dataset in R

Source: https://context7.com/sritchie73/ukbnmr/llms.txt

Loads the bundled toy dataset 'test_data' from the ukbnmr package and displays its structure and dimensions. This dataset is useful for testing package functions without real UK Biobank data.

```r
library(ukbnmr)

# Load test dataset
data(test_data)

# View structure - matches UK Biobank RAP export format
names(test_data)[1:10]
dim(test_data)
```

--------------------------------

### Access Sample Quality Control Information

Source: https://context7.com/sritchie73/ukbnmr/llms.txt

Loads the sample_qc_info data frame which contains metadata regarding the 18 sample processing and quality control fields in the UK Biobank NMR dataset.

```R
library(ukbnmr)

# Load sample QC information
data(sample_qc_info)
```

--------------------------------

### Extract sample-level processing and QC flags

Source: https://context7.com/sritchie73/ukbnmr/llms.txt

Retrieves metadata regarding sample processing, including shipment plates, well positions, and spectrometer identifiers. This is critical for identifying batch effects and sample-level quality issues.

```R
library(ukbnmr)
library(data.table)

# Load UK Biobank data
exported <- fread("path/to/exported_ukbiobank_phenotype_data.csv")

# Extract sample processing information
sample_qc_flags <- extract_sample_qc_flags(exported)

# Save sample QC information
fwrite(sample_qc_flags, file="path/to/nmr_sample_qc_flags.csv")
```

--------------------------------

### Access NMR Biomarker Metadata

Source: https://context7.com/sritchie73/ukbnmr/llms.txt

Demonstrates how to load and query the nmr_info data frame to retrieve biomarker descriptions, units, types, UKB field IDs, and calculation formulas.

```R
library(ukbnmr)

# Load biomarker information
data(nmr_info)

# Filter to specific types
non_derived <- nmr_info[Type == "Non-derived"]
composite <- nmr_info[Type == "Composite"]

# Get UK Biobank field ID for a specific biomarker
nmr_info[Biomarker == "Albumin", .(Biomarker, UKB.Field.ID, Description)]

# View formulas for derived biomarkers
nmr_info[Type == "Ratio", .(Biomarker, Simplified.Formula)][1:5]
```

--------------------------------

### Extract Biomarkers and QC Flags with Test Data in R

Source: https://context7.com/sritchie73/ukbnmr/llms.txt

Demonstrates how to use ukbnmr package functions to extract biomarkers, biomarker QC flags, and sample QC flags from the test dataset. These functions prepare the data for further analysis.

```r
# Test package functions with toy data
nmr <- extract_biomarkers(test_data)
qc_flags <- extract_biomarker_qc_flags(test_data)
sample_info <- extract_sample_qc_flags(test_data)
```

--------------------------------

### Process Test Data with Full QC Pipeline in R

Source: https://context7.com/sritchie73/ukbnmr/llms.txt

Applies the full technical variation removal pipeline from the ukbnmr package to the test dataset. This function is used to clean the data by accounting for technical sources of variation.

```r
# Full QC pipeline also works on test data
processed <- remove_technical_variation(test_data)
```

--------------------------------

### Recompute derived biomarkers and QC flags in R

Source: https://github.com/sritchie73/ukbnmr/blob/main/README.md

Demonstrates how to remove technical variation from UK Biobank NMR data and subsequently recompute composite biomarkers and ratios. This process ensures data consistency after adjusting for biological covariates.

```R
library(ukbnmr)
library(data.table)
library(MASS)

# Load exported field data
exported <- fread("path/to/exported_ukbiobank_phenotype_data.csv")

# Remove technical variation
processed <- remove_technical_variation(exported)
tech_qc <- processed$biomarkers

# Write out results
fwrite(tech_qc, file="path/to/nmr_biomarker_data.csv")
fwrite(processed$biomarker_qc_flags, file="path/to/nmr_biomarker_qc_flags.csv")
fwrite(processed$sample_processing, file="path/to/nmr_sample_qc_flags.csv")
fwrite(processed$log_offset, file="path/to/nmr_biomarker_log_offset.csv")
fwrite(processed$outlier_plate_detection, file="path/to/outlier_plate_info.csv")
```

--------------------------------

### Compute Extended Ratio Quality Control Flags

Source: https://context7.com/sritchie73/ukbnmr/llms.txt

Generates specific QC flags for the 76 extended biomarker ratios. This function maps quality issues from source biomarkers to the computed ratios.

```R
library(ukbnmr)
library(data.table)

# Compute QC flags for extended ratios
extended_qc_flags <- compute_extended_ratio_qc_flags(exported)

# Save QC flags
fwrite(extended_qc_flags, file = "path/to/extended_ratio_qc_flags.csv")
```

--------------------------------

### Compute extended biomarker ratios in R

Source: https://context7.com/sritchie73/ukbnmr/llms.txt

The compute_extended_ratios function calculates 76 additional lipid and fatty acid ratios beyond the standard Nightingale platform output, preserving original biomarker data in the process.

```R
library(ukbnmr)
library(data.table)

exported <- fread("path/to/exported_ukbiobank_phenotype_data.csv")
nmr_extended <- compute_extended_ratios(exported)
```

--------------------------------

### Extract NMR biomarkers

Source: https://context7.com/sritchie73/ukbnmr/llms.txt

Converts raw UK Biobank field data into an analysis-ready data frame with descriptive biomarker names. It maps cryptic field identifiers to meaningful names and organizes data by participant ID and visit index.

```R
library(ukbnmr)
library(data.table)

# Load exported field data from UK Biobank Research Analysis Platform
exported <- fread("path/to/exported_ukbiobank_phenotype_data.csv")

# Extract biomarkers with descriptive column names
nmr <- extract_biomarkers(exported)

# Save for later use
fwrite(nmr, file="path/to/nmr_biomarker_data.csv")
```

--------------------------------

### Calculate and Save Extended NMR Biomarker Ratios

Source: https://context7.com/sritchie73/ukbnmr/llms.txt

Identifies non-Nightingale extended ratios from the nmr_info dataset and extracts them from the main biomarker table. The resulting subset is saved to a CSV file for downstream analysis.

```R
library(ukbnmr)
library(data.table)

# Extended ratios are marked in nmr_info where Nightingale == FALSE
extended_ratio_names <- ukbnmr::nmr_info[Type == "Ratio" & Nightingale == FALSE, Biomarker]

# Check values of extended ratios
head(nmr_extended[, .SD, .SDcols = c("eid", "visit_index", extended_ratio_names[1:5])])

# Save extended biomarker data
fwrite(nmr_extended, file = "path/to/nmr_extended_ratios.csv")
```

--------------------------------

### Adjust NMR Biomarkers for Covariates in R

Source: https://github.com/sritchie73/ukbnmr/blob/main/README.md

This process involves merging phenotypic data with NMR biomarkers, applying log-transformation with offsets, performing robust linear regression to remove covariate effects, and recomputing derived biomarkers. It relies on the data.table and MASS packages to handle large-scale biological datasets efficiently.

```R
library(data.table)
library(MASS)

# Prepare covariates
baseline_covar <- exported[, .(eid, age=p21003_i0, sex=p31_i0, bmi=p21001_i0)]
repeat_covar <- exported[, .(eid, age=p21003_i1, sex=p31_i1, bmi=p21001_i1)]
covar <- rbind(idcol="visit_index", "0"=baseline_covar, "1"=repeat_covar)
covar[, visit_index := as.integer(visit_index)]

# Merge and transform
non_derived <- ukbnmr::nmr_info[Type == "Non-derived", Biomarker]
bio_qc <- tech_qc[, .SD, .SDcols=c("eid", "visit_index", intersect(non_derived, names(nmr)))]
bio_qc <- covar[bio_qc, on = .(eid, visit_index), nomatch=0]
bio_qc <- melt(bio_qc, id.vars=c("eid", "visit_index", "age", "sex", "bmi"), variable.name="biomarker")

# Adjust using robust linear regression
bio_qc[, adjusted := rlm(biomarker ~ age + factor(sex) + log(bmi))$residuals, by=biomarker]
bio_qc[, adjusted := exp(adjusted + as.vector(coef(rlm(value ~ 1)))[1]), by=biomarker]

# Recompute derived biomarkers and save
bio_qc_final <- recompute_derived_biomarkers(dcast(bio_qc, eid + visit_index ~ biomarker, value.var="adjusted"))
fwrite(bio_qc_final, file="path/to/nmr_biomarkers_adjusted_for_covariates.csv")
```

--------------------------------

### Extract UK Biobank NMR Sample Quality Control Flags

Source: https://github.com/sritchie73/ukbnmr/blob/main/README.md

Extracts NMR sample processing and quality control flags from exported UK Biobank field data. It returns a data.frame with one column per flag type, indicating flags for each participant observation. Requires the 'data.table' package for efficient CSV reading.

```r
library(ukbnmr)
library(data.table) # for fast reading and writing of csv files using fread() and fwrite()

# Load exported field data saved by the Table Exporter tool on the RAP
exported <- fread("path/to/exported_ukbiobank_phenotype_data.csv")

sample_qc_flags <- extract_sample_qc_flags(exported)

fwrite(sample_qc_flags, file="path/to/nmr_sample_qc_flags.csv")
```

--------------------------------

### Extract biomarker quality control flags

Source: https://context7.com/sritchie73/ukbnmr/llms.txt

Extracts QC flags for individual biomarker measurements to identify issues like values below the limit of quantification or technical errors. Flags are returned in a data frame aligned with participant IDs.

```R
library(ukbnmr)
library(data.table)

# Load UK Biobank data
exported <- fread("path/to/exported_ukbiobank_phenotype_data.csv")

# Extract biomarker QC flags
biomarker_qc_flags <- extract_biomarker_qc_flags(exported)

# Save QC flags
fwrite(biomarker_qc_flags, file="path/to/nmr_biomarker_qc_flags.csv")
```

--------------------------------

### Extract UK Biobank NMR Biomarker Quality Control Flags

Source: https://github.com/sritchie73/ukbnmr/blob/main/README.md

Extracts quality control flags for NMR biomarker measurements from exported UK Biobank field data. It returns a data.frame with one column per biomarker, indicating QC flags for each participant observation. Observations with no flags are NA, and multiple flags are semicolon-separated. Requires the 'data.table' package for efficient CSV reading.

```r
library(ukbnmr)
library(data.table) # for fast reading and writing of csv files using fread() and fwrite()

# Load exported field data saved by the Table Exporter tool on the RAP
exported <- fread("path/to/exported_ukbiobank_phenotype_data.csv")

biomarker_qc_flags <- extract_biomarker_qc_flags(exported)

fwrite(biomarker_qc_flags, file="path/to/nmr_biomarker_qc_flags.csv")
```

--------------------------------

### Remove Technical Variation and Save NMR Data (R)

Source: https://github.com/sritchie73/ukbnmr/blob/main/README.md

This R code snippet demonstrates how to use the `remove_technical_variation` function to process UK Biobank NMR data and save the processed biomarker data, QC flags, and sample processing information to CSV files. It requires the `ukbnmr` and `data.table` libraries and assumes the input data is loaded from a CSV file.

```r
library(ukbnmr)
library(data.table) # for fast reading and writing of csv files using fread() and fwrite()

# Load exported field data saved by the Table Exporter tool on the RAP
exported <- fread("path/to/exported_ukbiobank_phenotype_data.csv")

processed <- remove_technical_variation(exported) 

fwrite(processed$biomarkers, file="path/to/nmr_biomarker_data.csv")
fwrite(processed$biomarker_qc_flags, file="path/to/nmr_biomarker_qc_flags.csv")
fwrite(processed$sample_processing, file="path/to/nmr_sample_qc_flags.csv")
fwrite(processed$log_offset, file="path/to/nmr_biomarker_log_offset.csv")
fwrite(processed$outlier_plate_detection, file="path/to/outlier_plate_info.csv")
```

--------------------------------

### Recompute Derived Biomarker Quality Control Flags

Source: https://context7.com/sritchie73/ukbnmr/llms.txt

Aggregates QC flags from underlying biomarkers for all composite and derived biomarkers. This ensures that if a component biomarker has a quality issue, it is reflected in the derived biomarker's flag.

```R
library(ukbnmr)
library(data.table)

# Load UK Biobank data
exported <- fread("path/to/exported_ukbiobank_phenotype_data.csv")

# Compute aggregated QC flags for all derived biomarkers
biomarker_qc_flags <- recompute_derived_biomarker_qc_flags(exported)

# Save aggregated QC flags
fwrite(biomarker_qc_flags, file = "path/to/biomarker_qc_flags_aggregated.csv")
```

--------------------------------

### Recompute derived biomarkers in R

Source: https://context7.com/sritchie73/ukbnmr/llms.txt

The recompute_derived_biomarkers function recalculates composite and ratio-based biomarkers from non-derived inputs. This is crucial for maintaining data consistency after applying biological covariate adjustments.

```R
library(ukbnmr)
library(data.table)

# Assuming bio_qc contains adjusted non-derived biomarkers
bio_qc <- recompute_derived_biomarkers(bio_qc)

fwrite(bio_qc, file = "path/to/nmr_biomarkers_adjusted.csv")
```

--------------------------------

### Remove technical variation from NMR data using R

Source: https://context7.com/sritchie73/ukbnmr/llms.txt

The remove_technical_variation function cleans biomarker data by adjusting for spectrometer drift, plate effects, and sample degradation. It returns a list containing corrected concentrations, QC flags, and diagnostic information.

```R
library(ukbnmr)
library(data.table)

exported <- fread("path/to/exported_ukbiobank_phenotype_data.csv")
processed <- remove_technical_variation(exported)

nmr_corrected <- processed$biomarkers
qc_flags <- processed$biomarker_qc_flags

# Save results
fwrite(processed$biomarkers, file="path/to/nmr_biomarker_data_qc.csv")
```
Tokens:
Raw