Skip to the content.

Reproducible Python demos using synthetic data shaped like biological and chemical experiments. Each piece links to code under demos/ in the GitHub repository. Figures are checked in under each demo’s outputs/ folder.

Differential expression (toy)

Gene-level summaries across two conditions; volcano plot with Benjamini–Hochberg FDR.

Compound similarity (no RDKit)

Synthetic fingerprints and physicochemical properties; Tanimoto similarity and PCA landscape.

Dose–response / viability

Hill-style synthetic curves; fitted potency and uncertainty for prediction-style readouts.

Residue contact map

Synthetic Cα distances; binary contact map at a distance threshold.

Generative sequences (PWM + latent)

Motif sampling with strength modulated by a 2D latent vector; scores and interpolation.

Multimodal biological samples

Expression, imaging-style features, and clinical covariates; fused view with CCA.

Dependencies: demos/requirements.txt. Run python3 data/generate.py then python3 src/run.py in each demo folder.