Article: Tau protein transmission simulation modelling in Alzheimer's disease integrated with neuro-symbolic learning Journal: International Journal of Simulation and Process Modelling (IJSPM) 2026 Vol.23 No.6 pp.1 - 12 Abstract: Transneuronal propagation of Tau in Alzheimer's disease is a central mechanism of disease progression. We propose a computational framework incorporating neuro symbolic learning to model this multi-scale process. This model combines graph neural networks with symbolic rules encoding biological priors (e.g., prion-like propagation, metabolic activity promotion), and constructs a bridge from 'cognitive load theory' to 'computable pathological model'. Experiments based on the Alzheimer's disease neuroimaging initiative real dataset (n = 428) show that NSTP-Net has the root mean square error (0.135) significantly lower than the current advanced methods in predicting the Tau distribution over the next 18 months, and the performance improvement reaches 22% (p < 0.001). The model also had high accuracy in predicting Mild Cognitive Impairment to Alzheimer's disease conversion. This study provides an interpretable new tool for understanding the mechanism of Tau propagation and demonstrates important clinical potential for individualised prognosis prediction. Inderscience Publishers - linking academia, business and industry through research

Title: Tau protein transmission simulation modelling in Alzheimer's disease integrated with neuro-symbolic learning

Authors: Mengke Huo; Yajing Chen; Huiqin Wang

Addresses: School of Health Medicine, Zhengzhou Health College, Zhengzhou, 450000, China ' School of Health Medicine, Zhengzhou Health College, Zhengzhou, 450000, China ' School of Health Medicine, Zhengzhou Health College, Zhengzhou, 450000, China

Abstract: Transneuronal propagation of Tau in Alzheimer's disease is a central mechanism of disease progression. We propose a computational framework incorporating neuro symbolic learning to model this multi-scale process. This model combines graph neural networks with symbolic rules encoding biological priors (e.g., prion-like propagation, metabolic activity promotion), and constructs a bridge from 'cognitive load theory' to 'computable pathological model'. Experiments based on the Alzheimer's disease neuroimaging initiative real dataset (n = 428) show that NSTP-Net has the root mean square error (0.135) significantly lower than the current advanced methods in predicting the Tau distribution over the next 18 months, and the performance improvement reaches 22% (p < 0.001). The model also had high accuracy in predicting Mild Cognitive Impairment to Alzheimer's disease conversion. This study provides an interpretable new tool for understanding the mechanism of Tau propagation and demonstrates important clinical potential for individualised prognosis prediction.

Keywords: Alzheimer's disease; Tau protein transmission; neuro symbolic learning; graph neural network; computational pathological model.

DOI: 10.1504/IJSPM.2026.153267

International Journal of Simulation and Process Modelling, 2026 Vol.23 No.6, pp.1 - 12

Received: 17 Dec 2025
Accepted: 02 Feb 2026
Published online: 29 Apr 2026 *

Title: Tau protein transmission simulation modelling in Alzheimer's disease integrated with neuro-symbolic learning

Keep up-to-date