1 of 1

Introduction

Highlights

Dependency Documentation

​

  • MetabolonTM is a targeted metabolomics assay that measures metabololite levels nested within super-pathways.

​

  • We have developed a flexible analysis pipeline which utilizes the best practices in reproducibility for the quality control and analysis of Metabolon data.

​

​

​

​

​

​

​

​

​

​

​

  • The pipeline is open source and shared through GitHub
  • Each analysis step stands on its own with minimal dependence on other steps
  • A workflow diagram that clearly demonstrates the input data and the output for each analysis step
  • A vignette that demonstrates each step of the analysis pipeline
  • Written generally to handle differing experimental designs and conditions
  • Environment Reproducibility using renv
  • Shiny App with a user-friendly interface

​

​

​

�

GitHub Documentation

MetabolomicsPipeline: An analysis pipeline for metabolomics research

Joel Parker¹, Dr. Bonnie LaFleur2,3, Lisa White², David Bradford4, Syed Zaidi5, Rudramani Pokhrel6

Mel and Enid Zuckerman College of Public Health¹, BIO5 Institute², R. Ken Coit College of Pharmacy³, Center of Innovation and Brain Science4, College of Medicine5, University of Arizona Cancer Center6

​

Results and Figures

Figure 6: The pipeline is available through an R shiny app which allows users to upload their data or use the example data as an exemplar. We also allow the option to include metadata that might not be included in Metabolon-generated reports.

Acknowledgements

References

We would like to thank David Bradford for providing context around the exemplar data used for this pipeline. Lisa White built the Shiny App and along with Syed Zaidi and Rudramani Pokhrel wrote the code that served as a template for the workflow. We would like to thank Dr. Bonnie LaFleur who served as a mentor and PI for this project. This work has been partially supported by the University of Arizona Aegis Consortium, National Institutes on Aging (AG065169), and the National Cancer Institute (2P01CA229112-06).�

Shiny App

Figure 1: The pipeline can be utilized through an R shiny wrapper or run locally.

1. Separate files for each step

4. Vignette to demonstrate pipeline

5. Vignette to demonstrate pipeline

Figure 2: Github repository contains:

​

  • Folder for each of the .R files containing each step of the pipeline
  • Folder for the raw and processed data
  • Folder for help files for each function in our package
  • Vignette to demonstrate the pipeline using the example data
  • Lock file documenting package versions for environment reproducibility

https://rstudio.github.io/renv/articles/renv.html

Figure 3: We use the renv package to store R environment information to easily update upgrade/downgrade local package versions each time the pipeline is used.

Workflow Diagram

Figure 4: We provide a workflow diagram that labels the input data and the output for each step of the analysis. We also highlight the dependencies between each of the steps in the workflow.

R Package

Figure 5: Creating an R package for internal functions lets us create help files for each function that are easily accessible for the user.

Table 1: The pipeline is generalized to handle differing experimental designs and conditions. Results from the pipeline are presented as publishable tables and figures.

QR Codes

GitHub

Shiny App

Ushey K, Wickham H (2024). renv: Project Environments. R package version 1.0.5, https://github.com/rstudio/renv, https://rstudio.github.io/renv/.

​

Wickham, H., & Bryan, J. (2023). R packages. " O'Reilly Media, Inc.".

​

​

​

​