Skip to content

Scripts for analysis of metabolomics data relating to nutrition, health and disease in the EPIC study.

Notifications You must be signed in to change notification settings

JoeRothwell/EPIC-metabolomics

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

66 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

EPIC metabolomics

This repository puts together scripts that contain various different workflows for untargeted metabolomics data in the EPIC study. Analyses are for alcohol, coffee and acylcarnitine projects.

Manhattan plot showing results of a metabolome wide association study (MWAS) on coffee intake. Each point represents a spectral feature generated by the mass spectrometer after injection of serum samples and separation of components.

Description of files

alcohol_study_untarg.R runs two workflows for discovery of alcohol-related spectral features from untargeted metabolomics analyses. In the first, CS and HCC datasets are feature matched (with m/z and RT tolerances; fuzzyjoin package) and the remaining features subject to partial correlation with alcohol intake. In the second, only features from the CS study associated with alcohol intake are matched with all features from the HCC study, and the matched HCC features then tested for correlations with alcohol intake. Note: functions from Intake_correlation_new.R are required and must be run first.

Intake_cor_functions.R contains only the functions needed to run alcohol_study_untarg.R.

Intake_correlation_new.R Functions that perform a metabolome-wide association study for food intake in the EPIC cross-sectional study. Partial correlations are computed between spectral feature intensities and food intakes or lifestyle factors of interest, adjusting for confounders. The output is a data frame consisting of one row per feature, with other information such as median intensity and number of detections also extracted from the feature table. This output is a basis for compound identification, and two further functions use it to generate a correlation matrix or a Manhattan plot.

Feature_matching.R matches features from different untargeted metabolomics feature tables given mass and retention time tolerances. (For alcohol biomarker study)

CS_descriptives_coffee.R analyses coffee intake data in EPIC and extracts coffee intake data from the whole dataset.

Coffee_biomarkers_CS.R and Coffee_biomarkers_HCC.R calculate associations between known coffee intake biomarkers and factors of interest or HCC status in the EPIC cross-sectional and HCC studies respectively.

Acylcarnitines.R is an analysis of acylcarnitine plasma intensities extracted from untargeted data. Compounds correlations with food intake are tested and a heatmap plotted. Blood acylcarnitine intensies are also tested for associations with WCRF score.

About

Scripts for analysis of metabolomics data relating to nutrition, health and disease in the EPIC study.

Topics

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages