This project assesses whether knowing the abundance of selected gut bacteria well-documented in IBD research reduces uncertainty in Crohn’s disease status. Because mutual information captures both linear and nonlinear associations without assuming a specific parametric model, it was chosen to quantify dependence between bacterial abundance and disease status. Using shotgun metagenomics–derived microbial relative abundance of the 9 selected bacteria, I computed mutual information with 10,000 permutations to assess significance, and applied false discovery rate (FDR) control to account for multiple hypothesis testing.
- Source: GMrepo (Project PRJEB42155)
- Samples: 132 unique tube_ids (18 Healthy, 114 Crohn’s Disease)
- Attributes: tube_id, diagnosis (Crohn’s disease, healthy), shotgun metagenomics–derived microbial relative abundance of the 9 selected bacteria
- Filtered data to obtain healthy controls vs Crohn’s disease
- Binned relative abundance into 3 bins (zero, low, high)
- Calculated the mutual information between each bacterium and disease status
- Ran permutation tests (10,000 permutations) to assess significance
- Corrected for multiple hypothesis testing with FDR correction
Mutual information values were low across all taxa, indicating substantial overlap between Crohn’s disease and healthy samples, with only a subset remaining significant after FDR correction.
- Python 3.10+
- pandas >= 2.0
- matplotlib >= 3.9
data/raw/– original input data (TSV files)cleaned/– analysis-ready datasetjoined_MI_ready.csv
scripts/clean_and_join.py– data cleaning, normalization, and joiningbinning_and_MI.py– binning, MI computation, permutation testing, FDR correction
results/figures/– plots and visual summariesMI_dotplot.pngboxplot_Faecalibacterium.pngboxplot_Faecalibacterium_prausnitzii.pngboxplot_Escherichia_coli.png
tables/– machine-readable result tablesMI_results.csv
README.mdLICENSE
MIT
Course: CSDS 313 – Intro to Data Analysis, Instructor Dr. Mehmet Koyuturk