
ABSTRACT This study explores the application of biological invasion models in epidemiology by examining the spatial relationship between dengue fever transmission and the suitable habitats of Aedes aegypti and Aedes albopictus in China. By integrating invasion biology and epidemiological modelling perspectives, the study applies the MaxEnt model to predict mosquito suitable habitats using occurrence records and environmental variables, and further combines the spatial Durbin model to analyses the direct and indirect spatial effects of mosquito distribution on dengue incidence. Results reveal significant spatial correlations between mosquito habitats and dengue outbreak regions. A. aegypti exhibits stronger direct effects on local dengue transmission, suggesting higher infectivity within suitable habitats, whereas A. albopictus demonstrates stronger spatial spillover effects, indicating greater diffusion capacity across regions. The findings highlight the value of integrating species distribution modelling with spatial econometric analysis to improve understanding of vector‐borne disease diffusion and support targeted prevention and control strategies from an interdisciplinary systems perspective.
This study investigates the transmission dynamics of streptococcal disease using a stochastic delay differential equation (SDDE) framework. The population is divided into four compartments: susceptible individuals , infected individuals , carriers of streptococcus and individuals who progress to the post-infectious complication of acute rheumatic fever (ARF), representing severe disease outcomes rather than an infectious transmission class, . The incorporated discrete time delay represents the effective latency associated with bacterial colonisation, disease incubation and the early host immune response prior to disease progression. The deterministic delayed model is first analysed to establish fundamental dynamical properties, including positivity, boundedness and the existence and uniqueness of solutions. The equilibrium states are derived and their local and global stability are examined in relation to the basic reproduction number, , which serves as the epidemiological threshold determining whether the disease dies out or persists within the population. A stochastic delayed version of the model is then formulated, and conditions ensuring the existence of a unique global positive solution are obtained. Furthermore, a dynamically consistent stochastic nonstandard finite difference (NSFD) scheme is constructed by combining the nonstandard finite difference methodology with an Euler-Maruyama discretisation of the stochastic perturbation terms to preserve the qualitative dynamics of the continuous SDDE model. The proposed scheme preserves essential dynamical properties of the system and is shown to converge to the corresponding equilibrium states without restrictions on the step size. Numerical simulations are carried out to compare the proposed scheme with several standard stochastic numerical methods. The results demonstrate that the nonstandard finite difference scheme maintains positivity and boundedness of the solutions and closely follows the mean trajectory of the deterministic model, whereas the standard schemes may violate these properties. The proposed framework therefore provides a reliable and dynamically consistent computational tool for investigating the influence of stochastic perturbations and delayed disease progression on streptococcal transmission dynamics. Finally, the influence of stochasticity and time delay on the dynamics of susceptible and infected populations is illustrated through numerical experiments.
The brain network has proven to be an effective tool for recognising neurological disorders. However, current brain network modelling algorithms often rely on numerous assumptions when evaluating the interaction relationships between brain regions, which are frequently shown to be inaccurate. For instance, some studies assume that interactions among brain regions are linear. Additionally, some research posits that other brain regions do not influence the evaluation outcomes when assessing causal effects among paired brain regions. To address this challenge, we propose an innovative method for modelling brain networks that estimates interaction relationships among brain regions from a causal perspective. Notably, during the evaluation process, this method comprehensively considers all relevant brain regions rather than examining individual relationships in isolation. To validate the efficacy of our approach, we conducted extensive experimental verifications using publicly available datasets. The experimental results indicate that our proposed brain network modelling method not only demonstrates superior performance in the dementia identification task but also successfully identifies several new biomarkers, providing novel perspectives and insights for dementia research.
ABSTRACT Multiple studies have demonstrated a significant correlation between histone modifications and tumour initiation and progression. However, it remains unclear whether the histone modification profile contributes to the classification of patients with gastric cancer (GC). After analysing transcriptomic data from 76 histone modification regulators (HMR), we identified four distinct patterns in 838 GC specimens. These patterns demonstrated associations with different prognoses, treatment responses, biological pathways, and tumour microenvironment characteristics. Therefore, we used the Least Absolute Shrinkage and Selection Operator (LASSO) regression and multivariate Cox regression analysis to explore GC's histone modification regulator characteristics and established a 10‐gene scoring scheme named “Histone Modification Regulator‐based Risk Score (HR_score)” to cluster tumours into high‐ and low‐risk subtypes. Patients with high scores are characterised by increased stromal cell infiltration and exhibit the activation of innate immunity. These observations may be linked to the immune‐exclusion phenotype and a poor prognosis. Further analysis revealed a direct correlation between higher HR_score and diminished response to adjuvant chemotherapy and immunotherapy treatments. In summary, this study indicates that HR_score can serve as a robust prognostic tool for GC, which can effectively perform patient risk stratification and predict treatment response, providing potential value for guiding individualised treatment strategies.
ABSTRACT Huntington's disease (HD) is a progressive neurodegenerative disease resulted from the expansion of CAG repeats in huntingtin (HTT) gene and characterised by motor impairment, psychiatric disorders and cognitive decline. Currently, there are no available disease‐modifying treatments. In this study, we performed an integrative analysis of HD datasets downloaded from the Gene Expression Omnibus (GEO) database using machine learning approaches, Mendelian randomisation (MR) and single cell RNA‐sequencing (scRNA‐seq) analysis to elucidate underlying disease mechanisms. The machine learning algorithms of LASSO, random forest and SVM‐RFE identified the decreased ADRA2A in HD. In addition, MR analysis revealed HD‐associated proteins such as LILRA4 and UBE2D1, associated with immune response and ubiquitination metabolism, respectively. scRNA‐seq analysis further identified the transcription factors of SOX4 and ETV1 through SCENIC analysis. Pseudotime trajectory analysis highlighted heparan sulphate sulfotransferase activity, involving genes CHST11, HS3ST3A1 and HS6ST3 that drive cell differentiation. Moreover, cell‐cell communication analysis demonstrated enhanced NRG1/3‐ERBB4 signalling pathway in HD. This study presents multiple potential targets for HD treatment but further experimental validation is required to confirm their functional relevance.