Learning interpretable network dynamics via universal neural symbolic regression

Jiao Hu, Jiaxu Cui, Bo Xiong Yang

Nature Communications 2025 article cited by 13

Source Open-access PDF More in Science & Research

importance 7/10 ai assisted Model Reduction and Neural Networks complex systems science stage: research

Problem

Discovering the governing equations of complex network dynamics from data, across fields like physics, ecology and epidemiology.

Why it matters

Uncovering hidden mechanisms of complex systems supports scientific understanding and decision-making.

What the AI did

A tool combining deep learning fitting with pre-trained symbolic regression to automatically learn symbolic patterns of state changes in complex networks.

Results

Outperformed state-of-the-art symbolic regression techniques on more than ten representative scenarios; no numeric metrics stated in abstract

Impact

Validated on 10+ scenarios and real-world systems including global epidemic data (abstract truncated).

Record

Venue	Nature Communications
Published	2025
Work type	article
Authors	Jiao Hu, Jiaxu Cui, Bo Xiong Yang
Citations	13 (OpenAlex, at import time)
References	75
Model or method	universal neural symbolic regression (deep learning + pre-trained symbolic regression)
Organization	not stated in source
Primary source	https://doi.org/10.1038/s41467-025-61575-7
Open access	gold
Keywords	neural symbolic regression, network dynamics, governing equations, interpretable AI, deep learning

Importance rating

7 / 10

significant — thousands to millions affected

Governing-equation discovery spans physics, biochemistry, ecology, epidemiology; abstract frames it as a fundamental scientific challenge assisting decision-making, but states no quantified scale.

Basis: inferred. Judged from the source text, which gives no explicit scale. Stated scale: not stated in source. Method and limits.

More in Science & Research

#	Paper	Year	Citations	Importance
1	Neural‐Symbolic Machine Learning for Retrosynthesis and Reaction Prediction Marwin Segler, Mark P. Waller	2017	549	7
2	Data-driven Chemical Reaction Prediction and Retrosynthesis Vishnu H Nair, Philippe Schwaller, Teodoro Laino	2019	32	7
3	Incorporating the image formation process into deep learning improves network performance Yue Li, Yijun P. Su, Min Yi Guo, Xiaofei Han, Jiamin Liu, Harshad D. V...	2022	123	6
4	Generative and reinforcement learning approaches for the automated de novo design of bioactive compounds Maria V. Korshunova, Niles Huang, Stephen J. Capuzzi, Dmytro S. Radche...	2022	101	6
5	CGBVS‐DNN: Prediction of Compound‐protein Interactions Based on Deep Learning Masatoshi Hamanaka, Kei Taneishi, Hiroaki Iwata, Jun Ye, Jianguo Pei,...	2016	70	6