Reproducibility package (simulation code and data) for the paper
An Open-Quantum-Systems Theory of Quantum-Biological Communication Channels Liang Dong, IEEE Transactions on Molecular, Biological, and Multi-Scale Communications (Special Feature on Quantum Biology Computing and Communications, Series IV). Manuscript TMBMC-QB-26-0001.
This repository contains the simulation scripts, raw inputs, simulation outputs, and verification suites needed to reproduce the numerical results reported in the paper. It contains code and data only; the manuscript and its figure sources are not included here. All code is pure Python (NumPy + SciPy).
.
├── code/ Python simulation and verification scripts
│ ├── common.py shared utilities (units, table I/O, binary capacity)
│ ├── exp_interference.py four-site dephasing-assisted interference network
│ ├── exp_fmo.py seven-site FMO complex (Adolphs--Renger)
│ ├── exp_radical_pair.py cryptochrome FAD/Trp radical pair
│ ├── exp_dna_tunnelling.py G--C proton-tunnelling mutation channel
│ ├── exp_ion_channel.py KcsA-inspired ion-throughput loop
│ ├── exp_fmo_pdb.py FMO couplings from PDB 3EOJ (structure check)
│ ├── exp_kcsa_pdb.py KcsA geometry from PDB 1BL8 (structure check)
│ ├── exp_classical_null.py quantum-vs-classical BIC/Delta-ell comparison
│ ├── exp_nai_sensitivity.py NAI reference-level sensitivity sweep
│ ├── audit_theorems.py 13 theorem-vs-simulation regression tests
│ └── verify_theory_support.py extended parameter-range theory checks
├── data/ raw inputs
│ ├── 3EOJ.pdb FMO trimer crystal structure (RCSB PDB)
│ └── 1BL8.pdb KcsA channel crystal structure (RCSB PDB)
├── results/ simulation outputs (*.dat tables, *.json summaries)
├── requirements.txt
├── run.sh one-command reproduction + verification
├── CITATION.cff
└── LICENSE
Python 3.10+ with the packages in requirements.txt:
pip install -r requirements.txtOnly numpy (>= 2.0) and scipy (>= 1.10) are needed.
Reproduce every result and run both verification suites with one command:
bash run.shThis regenerates results/, runs the 13-test theorem-vs-simulation
regression suite (audit_theorems.py), and runs the extended
parameter-range verification (verify_theory_support.py). It exits with a
non-zero status if any check fails.
To run an individual experiment (writing its outputs to results/):
cd code
python3 exp_fmo.py| Suite | What it checks |
|---|---|
code/audit_theorems.py |
13 regression tests pinning each theoretical claim at a representative operating point: Proposition 1 (data-processing / quantum-Fisher bound), Corollary 1 (Holevo bound), Proposition 2 (dark-state lock), Proposition 3 (serial-composition bottleneck), the WKB kinetic-isotope scaling, CPTP trace conservation, the Haberkorn yield budget, and the BIC/Delta-ell quantum-vs-classical discrimination. |
code/verify_theory_support.py |
The same claims stress-tested across parameter ranges: the Fisher bound across magnetic-field amplitude, the Holevo chain across dephasing rate, the dark-state lock and its linear onset, the first-passage-density identity, and the serial bottleneck over 500 random channel pairs. |
Each check computes the two sides of an inequality or identity independently, so a PASS is genuine evidence rather than a tautology.
| Script | Key outputs (results/) |
Figure in the paper |
|---|---|---|
exp_interference.py |
interference_curve.dat, interference_traj_*.dat |
Fig. 3 (Sec. V-A) |
exp_fmo.py |
fmo_dephasing_curve.dat, fmo_traj_*.dat |
Fig. 4 (Sec. V-B) |
exp_radical_pair.py |
rp_compass_theta.dat, rp_fisher_B.dat, rp_zeno_kT.dat |
Fig. 5 (Sec. V-C) |
exp_dna_tunnelling.py |
dna_temperature_curve.dat, dna_asymmetry.dat |
Fig. 6 (Sec. V-D) |
exp_ion_channel.py |
ion_dephasing.dat, ion_iv.dat, ion_impulse.dat |
Fig. 7 (Sec. V-E) |
exp_fmo_pdb.py |
fmo_J_compare.dat, fmo_pdb_dephasing.dat |
Fig. 8 (Sec. V-F); Fig. 4(a) overlay |
exp_kcsa_pdb.py |
kcsa_pdb_dephasing.dat, kcsa_distances.dat |
Sec. V-F (KcsA) |
exp_classical_null.py |
classical_null_interference.dat, classical_null_fmo.dat |
Fig. 9 (Sec. V-G) |
exp_nai_sensitivity.py |
nai_sensitivity_interference.dat, nai_sensitivity_fmo.dat |
Sec. III-D (NAI sensitivity) |
The figures themselves are typeset in the manuscript and are not part of this
repository; the table above maps each script and its data files to the
corresponding figure number in the paper. The two exp_*_pdb.py scripts
require the corresponding files in data/.
The two crystal structures in data/ are obtained from the RCSB Protein
Data Bank (https://www.rcsb.org) and are redistributed here for convenience:
- 3EOJ -- FMO antenna protein trimer (Tronrud, Wen, Gay, Blankenship, Photosynth. Res. 100, 79, 2009).
- 1BL8 -- KcsA potassium channel (Doyle et al., Science 280, 69, 1998).
All other inputs are numerical parameters drawn from the cited literature
(see the manuscript). The simulation outputs in results/ are produced
entirely by the scripts in code/.
In code/exp_dna_tunnelling.py the arguments V_F_eV and V_R_eV are
historical names and are not the forward and reverse activation barriers.
V_F_eV = 0.51 eV sets the barrier of the symmetric (untilted) double well
via lam = 4 * V_F_eV / a**4, and the pair fixes only the imposed asymmetry
Delta = V_F_eV - V_R_eV = 0.33 eV. Because the linear tilt also displaces the
extrema, the realized activation barriers of the potential actually
simulated are approximately 0.685 eV and 0.355 eV. These are computed by
realized_barriers() and reported as V_F_realized_eV / V_R_realized_eV in
results/dna_summary.json.
All published results were generated from the true potential and are unaffected by this naming; only the documentation and the reported barrier labels changed in v1.1.0.
- Code (
code/): MIT License (seeLICENSE). - Simulation outputs (
results/): Creative Commons Attribution 4.0 (CC-BY-4.0). - Crystal structures (
data/*.pdb): redistributed from the RCSB PDB under its terms of use.
If you use this software or data, please cite the paper:
L. Dong, "An Open-Quantum-Systems Theory of Quantum-Biological Communication Channels," IEEE Transactions on Molecular, Biological, and Multi-Scale Communications, vol. 12, pp. 858-874, 2026, doi: 10.1109/TMBMC.2026.3716243.
| DOI | Points to |
|---|---|
| 10.5281/zenodo.20646426 | concept DOI, always the newest version |
| 10.5281/zenodo.20646427 | v1.0.0 snapshot (2026-06-11) |
| 10.5281/zenodo.21880020 | v1.1.0 snapshot (2026-08-11) |
The article's data-availability footnote cites the v1.0.0 snapshot (10.5281/zenodo.20646427), which was the current release when the paper went to press. For the latest corrected version, use the concept DOI above, this repository, or the IEEE DataPort record 10.21227/s74a-hr21.
Machine-readable citation metadata for this repository is in CITATION.cff.
REPRODUCIBILITY.md documents a verification that these results are hardware-independent:
no GPU is used, there is no randomness in any published result, and every published
quantity survives unchanged at the reported precision across threading configurations and
across NumPy/SciPy major versions. One command reproduces everything: bash run.sh.
Liang Dong (liangdng@gmail.com). ORCID: 0000-0002-8585-1087.
Department of Electrical and Computer Engineering, Baylor University; and
Department of Radiology, The University of Texas Southwestern Medical Center.