BIOCLOCK

This repository contains the code used for the BIOCLOCK pilot study. The manuscript is currently being prepared for submission and will be available soon.

Scripts

This repository contains three main scripts:

• 1_preprocessing.R: This script performs quality control on the DNA methylation data, applies the 101 preprocessing pipelines, and estimates the epigenetic ages and blood leukocyte composition.
• 2_analysis.R: This script contains the statistical analyses perfomed on the epigenetic ages and other phenotypic variables.
• 3_figures.R: This script contains the code used to generate the figures in the manuscript.

It also contains four additional scripts:

• function_runPipelines.R: This script is used to run the 101 preprocessing pipelines and is adapted from this repository: https://github.com/anilpsori/_pipelines_and_biomarkers/tree/main.
• Biolearn_EpiAge.py: This script is used to calculate the Horvath1, Horvath2, GrimAge1 and GrimAge2 epigenetic ages using the Biolearn python package (https://github.com/bio-learn/biolearn).
• runcalcPCClock.R: This script is used to calculate epigenetic age according to the principal component clocks (PCHorvath1, PCHorvath2, PCGrimAge) and is adapted from this repository: https://github.com/MorganLevineLab/PC-Clocks.
• plottingFunctions.R: This script contains additional plotting fuctions used in 3_figures.R.

Directory structure

The scripts rely on the following directory structure:

.
├── Data
│   ├── Infinium
│   ├── Objects
│   └── Pheno
├── Images
│   ├── Paper
│   ├── PrePro
│   └── QC
├── Results
│   ├── Analysis
│   └── PrePro
└── Scripts
    └── Additional
        └── PC_Clocks

Environment Information

Platform: Linux-5.15.0-76-generic-x86_64-with-glibc2.35

OS: Linux (Ubuntu 22.04.5 LTS)

Machine: x86_64 x86_64-conda-linux-gnu (64-bit)

For reproducibility, the mamba/conda environments (one for R and one for Python) are available in the folder Environments. There is both a completely resolved environment and a minimal environment for each. In case of system incompatibilities when using the complete environment, use the minimal ones instead.

The following GitHub packages are installed in the R session using Bioconductor as they are not available in the conda package repository:

• illuminaHumanMethylationEPICv2manifest (GitHub release 0.99.1)
• illuminaHumanMethylationEPICv2anno.20a1.hg38 (GitHub release 0.99.0)

The PC-CLocks repository (https://github.com/MorganLevineLab/PC-Clocks) will need to be cloned into the Scripts/Additional folder and named "PC_Clocks" in order to calculate the PC clocks.

DOI

Zenodo DOI for this repository:

DOI: 10.5281/zenodo.14710286