Dataset [1] Download: https://www.science.org/doi/10.1126/science.1105809
Note: This dataset is an **interventional** dataset.

We selected data under two interventions: **cd3cd28** and **cd3cd28icam2**, because these interventions are exogenous and do not affect the causal graph structure or the conditional independence (CI) relationships.

Both files are provided in the `./Data Files/` directory.

We refer to the variable relationship diagram in **Figure 2** of [1], which reflects biological consensus and is treated as expert knowledge (ground truth).
Based on this, we selected:
- 50 conditionally **independent** (CI) pairs
- 50 conditionally **dependent** (NI) pairs

(Note: In [2], the conditioning sets for NI cases are generally large. We modified and extended some of them to better represent a broader range of scenarios.)

The pairs are saved in `CI.txt` and `NI.txt`, respectively. Each line is stored in the form X;Y;(Z).

Our experimental results can be found in:
- `Flow-Cytometry.ipynb`
- `Flow-Cytometry_ccit.ipynb` (run separately due to high computational cost of CCIT)

---

[1] Karen Sachs, Omar Perez, Dana Pe’er, Douglas A. Lauffenburger, and Garry P. Nolan.  
Causal protein-signaling networks derived from multiparameter single-cell data.
Science, 308(5721):523–529, 2005.

[2] Shuai Li, Yingjie Zhang, Hongtu Zhu, Christina Wang, Hai Shu, Ziqi Chen, Zhuoran Sun, and Yanfeng Yang.  
K-nearest-neighbor local sampling based conditional independence testing.
Advances in Neural Information Processing Systems, 36, 2024.
