Overview

Given a phylogenetic tree and gene cluster assignments (cls. IDs) from a real dataset, we:

1. Simulate genome evolution from a root genome across a phylogenetic tree.
2. Compute synteny blocks for all pairs of leaf genomes.
3. Compare distributions of simulated vs. real synteny block lengths using Wasserstein distance.
4. Optimize parameters using Bayesian optimization to minimize this distance.

---

Key Files

🔹 simulation.py

• Core simulation engine for genome evolution.
• Implements gain, loss, and inversion events along tree branches.
• Computes synteny blocks across all simulated leaf genome pairs.
• Provides run_simulation(...) to return a list of block lengths.

🔹 sim_real_comparison.py

• Loads the real phylogenetic tree and genome data.
• Loads true synteny block lengths (real_lengths.txt).
• Uses skopt to optimize gain/loss parameters via Bayesian optimization.
• Minimizes the Wasserstein distance between simulated and real block length distributions.

🔹 calc_real_data.py

• Parses real genome cluster assignments (atgc.cc.csv).
• Computes real synteny block lengths between leaf genomes using findSyntenyReal.
• Used to precompute and save the real_lengths.txt file.