Roadmap

[jiweiqi]

Below are some general thoughts about the Roadmap of developing CRNN, collaborations are welcomed.

Functionality and Scalability

• Reversible reactions
• Three-body reactions (pressure-dependence)
• Stiff reaction systems in combustion, biology, and climate
• Large system with more than 100 species, e.g., in combustion, biology, and climate
• QSSA: quasi-steady-state assumptions
• Reaction-diffusion system described by PDE
• Stochastic systems described by SDE
• Bayesian inference

Applications

• Homogenous combustion modeling
• Combustion with transport
• Electrochemistry in battery and fuel cell
• Multiphase reactions for fire safety, i.e., condensed phase and gas phase
• PK/PD: pharmacokinetic/pharmacodynamic modeling
• Single-cell sequence data
• High-throughput experiments
• Fire safety through better kinetic modeling
• Biomass production
• Control via better system identification
• Pattern formation

[jiweiqi]

QSSA

We already have a preliminary example in the yeast model. For combustion problems, the following paper might be useful

• Whitehouse, L. E., Tomlin, A. S., & Pilling, M. J. (2004). Systematic reduction of complex tropospheric chemical mechanisms, Part II: Lumping using a time-scale based approach. Atmospheric Chemistry and Physics, 4(7), 2057-2081.
• Porter, R., Fairweather, M., Griffiths, J. F., Hughes, K. J., & Tomlin, A. S. (2007). Simulated autoignition temperature and oxidation of cyclohexane using a QSSA reduced mechanism. In Proceedings of the 3rd European Combustion Meeting, Chania.

[jiweiqi]

Bayesian Inference

• Dandekar, Raj, Karen Chung, Vaibhav Dixit, Mohamed Tarek, Aslan Garcia-Valadez, Krishna Vishal Vemula, and Chris Rackauckas. "Bayesian neural ordinary differential equations." arXiv preprint arXiv:2012.07244 (2020).

[jiweiqi]

Stiff reaction systems

Reference papers

• Ji, Weiqi, Weilun Qiu, Zhiyu Shi, Shaowu Pan, and Sili Deng. "Stiff-PINN: Physics-Informed Neural Network for Stiff Chemical Kinetics." arXiv preprint arXiv:2011.04520 (2020).
• Kim, Suyong, Weiqi Ji, Sili Deng, and Christopher Rackauckas. "Stiff Neural Ordinary Differential Equations." arXiv preprint arXiv:2103.15341 (2021).

[jiweiqi]

Control via better system identification

Reference paper

• Bhadriraju, Bhavana, Mohammed Saad Faizan Bangi, Abhinav Narasingam, and Joseph Sang‐Il Kwon. "Operable adaptive sparse identification of systems: Application to chemical processes." AIChE Journal 66, no. 11 (2020): e16980.

[jiweiqi]

Pattern formation

• Zhao, Hongbo, Brian D. Storey, Richard D. Braatz, and Martin Z. Bazant. "Learning the physics of pattern formation from images." Physical review letters 124, no. 6 (2020): 060201.
• Sirimungkala, Atchara, Horst-Dieter Försterling, Vladimir Dlask, and Richard J. Field. "Bromination reactions important in the mechanism of the Belousov− Zhabotinsky system." The Journal of Physical Chemistry A 103, no. 8 (1999): 1038-1043.

Origins of Life: Introduction - Pattern Formation in Chemical Systems - Reaction Diffusion Systems

[jiweiqi]

Electrochemistry

• Kennedy, Gareth F., Jie Zhang, and Alan M. Bond. "Automatically identifying electrode reaction mechanisms using deep neural networks." Analytical chemistry 91.19 (2019): 12220-12227.

While the NN approach in this paper doesn't sound promising, the data generation process is quite a detail.

[jiweiqi]

Graph Neural Networks

• https://arxiv.org/ftp/physics/papers/0505/0505137.pdf
• Temkin, Oleg N., Andrew V. Zeigarnik, and D. G. Bonchev. Chemical reaction networks: a graph-theoretical approach. CRC Press, 1996.
• https://github.com/iMoonLab/THU-DeepHypergraph

Random questions

• can we learn the reactions through reconstructing the edges?
• how to learn hypergraph
• how does auto-differentiation come in
• do we need GNN-ODE?
• What's a deep random walk and how can we use it for learning GNN?