Link to Cell Reports Medicine article

INSTRUCTIONS

base_final.ipynb

Purpose:

This Jupyter Notebook performs the initial steps in the analysis of aberrant bowel movement frequencies (BMF) and their correlation with gut microbiome and organ function. It includes the following tasks:

1. Data Import and Preprocessing:

Imports raw data from Arivale snapshots and performs necessary cleaning, filtering, and transformation steps.

2. BMF Cohort Identification:

Defines criteria for BMF and identifies cohorts of individuals based on their bowel movement patterns.

3. Data Export:

Saves BMF cohort data into separate CSV files (e.g., asvs.csv) for subsequent analyses.

4. Descriptive Statistics:

Calculates and reports basic descriptive statistics for each BMF cohort, such as mean age, gender distribution, and other relevant metrics.

Input:

- arivale_snapshot:

Raw data from Arivale snapshots, containing relevant health and lifestyle information for each individual.

Output:

- [multi-omic].csv:

CSV files containing data for each identified BMF cohort.

Parameters:

- gender,age,BMI_CALC,vendor_dashboard,eGFR,CRP,A1C,LDL,PC[1-3],taxa_[taxonomic_classification],[metabolite_IDs],[clinical_chemistries],etc.:

List of parameters spanning the multi-omic data analyses needs and their BMF subcohorts (e.g., metabolomics).

Usage & Dependencies:

1. Set up environment & dependencies:

• Python
• pandas
• numpy
• seaborn

Ensure that the required Python packages (pandas, numpy, etc.) and dependencies are installed and loaded from the first few cells.

2. Specify input file:

Load the arivale_snapshot path within the notebook.

3. Run the notebook:

Execute the notebook cells sequentially or in desired chunks to complete the data import, preprocessing, cohort identification, and descriptive statistics calculation.

4. Review outputs:

Examine the generated CSV files and descriptive statistics summary.

metabolomics_eGFRanalysis_final.ipynb

Purpose:

This Jupyter Notebook investigates the relationship between BMF-associated metabolites and kidney function (estimated glomerular filtration rate - eGFR). It utilizes data from Arivale snapshots and the results of LIMMA regressions to perform an OLS regression analysis, outputting statistical summaries and plots.**

Input:

- Arivale snapshot data with relevant metabolomics and eGFR information.

- BMF-associated metabolites identified through LIMMA regressions.

Output:

- Statistical summaries of the OLS regression analysis (e.g., coefficients, p-values).

- Plots visualizing the relationships between metabolites and eGFR.

Usage & Dependencies:

- See base_final.ipynb for similar instructions and dependencies. Run the notebook to execute the analysis and generate the output files.

R Analysis Scripts (CORNCOB, LIMMA, POLR, etc.) and workspaces

Purpose:

This collection of R scripts performs statistical analyses on the preprocessed data generated by the Jupyter Notebooks. They leverage various R packages (e.g., bioconductor, tidyverse) to conduct regressions, including CORNCOB, LIMMA, and POLR. The scripts output graphical visualizations and summary statistics to aid in interpretation of the results.

Input:

- CSV files generated by the Jupyter Notebooks (e.g., BMF cohort data, metabolomics data, eGFR data).

Output:

- Graphical representations of the analysis results (e.g., plots, heatmaps).

- Summary statistics tables (e.g., regression coefficients, p-values).

Usage & Dependencies:

1. Set up environment & dependencies:

• R
• bioconductor
• tidyverse
• CORNCOB
• LIMMA
• polr

Ensure that the required R packages are installed and loaded.

2. Specify input files:

Adjust file paths in the scripts to match your directory structure.

3. Run the scripts:

Execute the scripts in R or RStudio to perform the analyses and generate the outputs.