Synthetic Crohn's disease microbiome-clinical profiles, ground-truth annotated

A collection of bio-synthetic dataset modeling the relationship between the gut and the disease. each patient "profile" is a unique 120-column vector of microbial and clinical data. the dataset is generated using gaussian copula expansion anchored to the microbiome atlas (n=71, id: 5), ensuring the synthetic data maintains the statistical covariance of real clinical cohorts. it includes a machine-readable ground_truth_key.json that defines the hidden biological rules (e.g., the exact correlation between f. prausnitzii and butyrate levels). designed specifically for training models to predict anti-tnf drug response rates.