Loading [MathJax]/jax/element/mml/optable/GeneralPunctuation.js
Research article

HRGCNLDA: Forecasting of lncRNA-disease association based on hierarchical refinement graph convolutional neural network


  • Long non-coding RNA (lncRNA) is considered to be a crucial regulator involved in various human biological processes, including the regulation of tumor immune checkpoint proteins. It has great potential as both a cancer biomolecular biomarker and therapeutic target. Nevertheless, conventional biological experimental techniques are both resource-intensive and laborious, making it essential to develop an accurate and efficient computational method to facilitate the discovery of potential links between lncRNAs and diseases. In this study, we proposed HRGCNLDA, a computational approach utilizing hierarchical refinement of graph convolutional neural networks for forecasting lncRNA-disease potential associations. This approach effectively addresses the over-smoothing problem that arises from stacking multiple layers of graph convolutional neural networks. Specifically, HRGCNLDA enhances the layer representation during message propagation and node updates, thereby amplifying the contribution of hidden layers that resemble the ego layer while reducing discrepancies. The results of the experiments showed that HRGCNLDA achieved the highest AUC-ROC (area under the receiver operating characteristic curve, AUC for short) and AUC-PR (area under the precision versus recall curve, AUPR for short) values compared to other methods. Finally, to further demonstrate the reliability and efficacy of our approach, we performed case studies on the case of three prevalent human diseases, namely, breast cancer, lung cancer and gastric cancer.

    Citation: Li Peng, Yujie Yang, Cheng Yang, Zejun Li, Ngai Cheong. HRGCNLDA: Forecasting of lncRNA-disease association based on hierarchical refinement graph convolutional neural network[J]. Mathematical Biosciences and Engineering, 2024, 21(4): 4814-4834. doi: 10.3934/mbe.2024212

    Related Papers:

    [1] Peter W. Bates, Jianing Chen, Mingji Zhang . Dynamics of ionic flows via Poisson-Nernst-Planck systems with local hard-sphere potentials: Competition between cations. Mathematical Biosciences and Engineering, 2020, 17(4): 3736-3766. doi: 10.3934/mbe.2020210
    [2] Yiwei Wang, Mingji Zhang . Finite ion size effects on I-V relations via Poisson-Nernst-Planck systems with two cations: A case study. Mathematical Biosciences and Engineering, 2024, 21(2): 1899-1916. doi: 10.3934/mbe.2024084
    [3] Dandan Fan, Dawei Li, Fangzheng Cheng, Guanghua Fu . Effects of congestion charging and subsidy policy on vehicle flow and revenue with user heterogeneity. Mathematical Biosciences and Engineering, 2023, 20(7): 12820-12842. doi: 10.3934/mbe.2023572
    [4] Tien-Wen Sung, Wei Li, Qiaoxin Liang, Chuanbo Hong, Qingjun Fang . Research on charging behavior of electric vehicles based on multiple objectives. Mathematical Biosciences and Engineering, 2023, 20(9): 15708-15736. doi: 10.3934/mbe.2023700
    [5] Chii-Dong Ho, Jr-Wei Tu, Hsuan Chang, Li-Pang Lin, Thiam Leng Chew . Optimizing thermal efficiencies of power-law fluids in double-pass concentric circular heat exchangers with sinusoidal wall fluxes. Mathematical Biosciences and Engineering, 2022, 19(9): 8648-8670. doi: 10.3934/mbe.2022401
    [6] Virginia González-Vélez, Amparo Gil, Iván Quesada . Minimal state models for ionic channels involved in glucagon secretion. Mathematical Biosciences and Engineering, 2010, 7(4): 793-807. doi: 10.3934/mbe.2010.7.793
    [7] Mette S. Olufsen, Ali Nadim . On deriving lumped models for blood flow and pressure in the systemic arteries. Mathematical Biosciences and Engineering, 2004, 1(1): 61-80. doi: 10.3934/mbe.2004.1.61
    [8] Guichen Lu, Zhengyi Lu . Permanence for two-species Lotka-Volterra cooperative systems with delays. Mathematical Biosciences and Engineering, 2008, 5(3): 477-484. doi: 10.3934/mbe.2008.5.477
    [9] Bindu Kumari, Chandrashekhar Sakode, Raghavendran Lakshminarayanan, Prasun K. Roy . Computational systems biology approach for permanent tumor elimination and normal tissue protection using negative biasing: Experimental validation in malignant melanoma as case study. Mathematical Biosciences and Engineering, 2023, 20(5): 9572-9606. doi: 10.3934/mbe.2023420
    [10] Suqing Lin, Zhengyi Lu . Permanence for two-species Lotka-Volterra systems with delays. Mathematical Biosciences and Engineering, 2006, 3(1): 137-144. doi: 10.3934/mbe.2006.3.137
  • Long non-coding RNA (lncRNA) is considered to be a crucial regulator involved in various human biological processes, including the regulation of tumor immune checkpoint proteins. It has great potential as both a cancer biomolecular biomarker and therapeutic target. Nevertheless, conventional biological experimental techniques are both resource-intensive and laborious, making it essential to develop an accurate and efficient computational method to facilitate the discovery of potential links between lncRNAs and diseases. In this study, we proposed HRGCNLDA, a computational approach utilizing hierarchical refinement of graph convolutional neural networks for forecasting lncRNA-disease potential associations. This approach effectively addresses the over-smoothing problem that arises from stacking multiple layers of graph convolutional neural networks. Specifically, HRGCNLDA enhances the layer representation during message propagation and node updates, thereby amplifying the contribution of hidden layers that resemble the ego layer while reducing discrepancies. The results of the experiments showed that HRGCNLDA achieved the highest AUC-ROC (area under the receiver operating characteristic curve, AUC for short) and AUC-PR (area under the precision versus recall curve, AUPR for short) values compared to other methods. Finally, to further demonstrate the reliability and efficacy of our approach, we performed case studies on the case of three prevalent human diseases, namely, breast cancer, lung cancer and gastric cancer.



    A rumor refers to unsubstantiated information, which has the potential to rapidly and extensively disseminate through various communication channels [1]. In the past, rumors were disseminated through oral communication as well as various forms of print media, radio broadcasts, and other mediums. Contemporary social networks can function as an appropriate platform for the rumor propagation, while also satisfying individuals' informational needs [2]. For example, individuals' excessive salt consumption can be attributed to the panic of the Fukushima nuclear accident-related rumor. There have been numerous rumors that emerged a range of adverse impacts on the world since the outbreak of COVID-19 [3]. The security of individuals and their belongings is significantly compromised due to these rumors. Therefore, it is imperative to grasp the rules of communicating with scientists and create effective methods to intervene, defend, and prevent rumors from spreading.

    Since rumor spreading is the analogy to the disease's transmission among individuals, many scholars employ similar rumor models. Early studies on rumor propagation focused a tiny community through individual interaction. According to rumor propagation, the crowd can be classified into three categories: ignorant, spreader, and stifler. In 1964, Daley and Kendall proposed the classic DK rumor-spreading model, which is the first cabin model in rumor propagation [4]. Furthermore, in 1973, Maki and Thompson proposed the MT model as an extension of the DK model [5]. According to the MT model, once interpersonal communication occurs, the former will no longer propagate. According to several scholars, the complexity of rumor propagation processes cannot be adequately captured by DK and MK models across various network settings, especially in the context of large-scale social networks. Zanette proposed a conceptual framework to simulate the dynamics of rumor propagation in small-world networks [6]. By integrating the MK model and the SIR model, Moreno et al. developed a novel rumor propagation model, demonstrating that network topologies can significantly influence the spread of rumors on scale-free networks [7]. To gain a deeper comprehension of rumor propagation, an increasing number of factors are being taken into account through the development of more rational models; for instance, hesitating mechanisms [8], inhibiting mechanisms, and attitude adjustment [9], etc. The innovations of these scholars, to a certain extent, promote rumor propagation to advance research under a variety of changing factors.

    Nowadays, given the multiple channels of rumor propagation, individuals can quickly obtain information on social media platforms, thereby complicating the regulation of rumor propagation. Due to the significant damage caused by rumors, several control measures aimed at minimizing losses have been proposed by scholars; for instance, optimal control strategies [10], time-dependent controller [11], event-triggered control [12], and so on. In reality, when rumors prevail on social media platforms, the media coverage disseminates information to debunk these rumors and intervene in individuals' behavior, thereby mitigating social panic and upholding social harmony. Cheng et al. incorporated a nonlinear factor into the contact rate by integrating optimal control and time delay [13]. Huo et al. introduced a nonlinear function to effectively capture the impact of misinformation and media reports on the spread of rumors [14]. Pan et al. investigated a novel rumor model that integrated media coverage and rumor refuters [15]. Further on, Guo et al. proposed two models considering media coverage and a refutation mechanism, respectively [16,17]. Numerous mathematical models are available to describe how media coverage affects rumor propagation. Therefore, our proposed model must account for the impact of media coverage on the dissemination of rumors among individuals. Evidently, the aforementioned models have predominantly explored the propagation mechanism of rumors in an integer order.

    However, the dynamic behavior of rumors is influenced by historical information. Fractional calculus, as an extension of integral calculus, provides innovative approaches for characterizing the memory effects, which is a generalization of integral calculus. Fractional order Caputo derivatives, extensively employed in various fields such as science, biology, engineering, computer science, chaos theory, and others, have witnessed significant advancements in recent decades [18,19,20,21,22,23,24,25]. The literature offers a wide range of models described by fractional derivatives, such as the SIR or SIRS model [26,27], the SEI model [28], the SEIR model [29,30,31], the SIQR model [32] and the SEIQRS model [33]. Rumor diffusion is similar to virus propagation and can be effectively modeled using fractional calculus. Singh proposed the dynamics of a rumor propagation in a social network, utilizing the Atangana-Baleanu derivative [34]. Ye et al. investigated a fractional-order reaction-diffusion rumor model in two linguistic contexts [35]. The impact of memory was investigated, and the findings clearly demonstrate that fractional calculus can be employed to more accurately characterize the process of rumor propagation.

    Revisiting the findings of the previous study, we establish an Ignorants-Spreaders-Calmness-Removers-Media Reports (ISCRM) rumor propagation model with media reports under the Caputo sense by incorporating the memory effect. The subsequent sections of this manuscript are organized as follows: In Section 2, we provide a concise overview of fundamental definitions pertaining to fractional calculus. In Section 3, we propose an enhanced fractional-order ISCRM rumor model. In Section 4, the local and global asymptotic stability of the equilibria is investigated. In Section 5, the optimal control problem is addressed. In Section 6, numerical simulation is examined. In Section 7, the paper culminates with the presentation of our findings.

    Various definitions exist regarding fractional derivatives, with the Caputo sense being provided as follows [36].

    Definition 2.1. For Riemann-Liouville fractional integration with respect to order α, the function f(t) is presented as

    t0Iαtf(t)=Dαtf(t)=1Γ(α)tt0f(τ)(tτ)1αdτ, (2.1)

    where αR+, t0 represents the initial time, and Γ(), which is established by Γ(z)=t0tz1etdt, represents the Gamma function.

    Definition 2.2. Define l1<αl, lN. The Riemann-Liouville fractional derivative with respect to order α is presented as [36]

    RLt0Dαtf(t)=dαf(t)dtα=1Γ(mα)dldtltt0f(τ)(tτ)αl+1dτ, (2.2)

    Remark 2.1. The combination of Definitions 2.1 and 2.2 yields RLt0Dαt(t0Iαtf(t))=f(t), giving the Riemann-Liouville fractional integrations and derivatives with respect to order α.

    Definition 2.3. The Caputo fractional derivative with respect to order α is presented as follows:

    Ct0Dαtf(t)={1Γ(lα)tt0fl(τ)(tτ)αl+1dτ, l1<α<l,dldtlf(t),α=l, (2.3)

    where l denotes the smallest integer number such that α<l.

    Remark 2.2. The equivalence between the definitions of Riemann-Liouville fractional integration and Caputo fractional integration are the same. Therefore, from Definitions 2.1 and 2.3, we obtain

    Ct0Dαt(t0Iαtf(t))=f(t), (2.4)

    giving the Caputo fractional integrations and derivatives with respect to order α.

    In the subsequent sections of this paper, the notation Dα represents the Riemann-Liouville fractional derivative.

    Lemma 2.1. Provide a definition for a function f(t) and a nonzero interval t(0,T), the condition |f(t)|F1, where F1>0, is a constant. Next, a positive constant is given L=F1TαΓ(1+α). Thus, the inequality 0Iαt1|f(t)|L is held for t(0,t1), t1T.

    Proof. The equality is evidently apparent:

    0Iαt1|f(t)|=1Γ(α)t10|f(τ)|(tτ)1αdτ. (2.5)

    By referring to Eq (2.1), which defines fractional integration, the value of 0Iαt1|f(t)| can be determined. Due to the condition |f(t)|F1, for at least one nonzero interval t(0,T), it acquires

    0Iαt1|f(t)|=1Γ(α)t10|f(τ)|(tτ)1αdτ1Γ(α)T0|f(τ)|(tτ)1αdτF1Γ(α)T01(tτ)1αdτ=F1TαΓ(1+α). (2.6)

    Therefore, the inequality 0Iαt1|f(t)|F1TαΓ(1+α) holds for f(t).

    Definition 2.4. Let F(τ) be the Laplace transform of the function f(t). Then, the Caputo fractional derivative Laplace transform is provided as [37]

    L[Dαf(t)]=ταF(τ)n1i=0f(i)(0)ταn1, n1<αn, nN. (2.7)

    Definition 2.5. The expressions for the Mittag-Leffler functions in both one-parameter and two-parameter forms are provided as [36]

    Eα(z)=k=0zkΓ(αk+1), Eα,l(z)=k=0zkΓ(αk+l), (2.8)

    where z,α,βC.

    Theorem 2.1. [36] Consider the autonomous linear fractional-order equation Dαx=Ax, x(0)=x0, where 0<α<1, xRn and ARn×n. The asymptotic stability of this system implies that

    |arg(eigA)|απ2. (2.9)

    In this scenario, each component of the state's decays toward 0 as tα.

    Theorem 2.2. [38] Let x=0 denote an equilibrium point for the equilibrium of the non-autonomous fractional-order equation

    Dαx(t)=f(x,t), (2.10)

    where f(x,t) satisfies the Lipschitz constant l>0 and α(0,1). Provided that it constitutes a Lyapunov function V(t,x(t)) and class-K functions α1, α2 and α3 satisfying the condition

    α1(

    where \beta \in \left({0, 1} \right) , the equilibrium point of Eq (2.10) is asymptotic stable.

    The proposed framework in this section is an Ignorants-Spreaders-Calmness-Removers-Media Reports (ISCRM). All of the crowd are divided into four groups: Ignorants are those who are unfamiliar with the rumors but are willing to believe them, denoted by I\left(t \right) ; Spreaders are those who acquire the rumors and spread them, denoted by S\left(t \right) ; Calmness are those who calm down before they refrain from disseminating rumors [39], denoted by C\left(t \right) ; Removers are those who are aware of the rumors and either refute them or have become disinterested in rumor, denoted by R\left(t \right) . Since rumor propagation varies over time, media reports will undergo changes over time. We employ a quantization method to measure the impact of media reports, denoted by M\left(t \right).

    The rules for rumor propagation are as follows:

    (ⅰ) When an ignorant contacts a spreader, it will be affected and become a spreader with the exposure rate of \beta /\left({1 + bS} \right) , which reflects the reduced amount of contact rate due to the scale of the rumor propagation. \beta denotes the rumor contact rate and b is the measure of inhibition.

    (ⅱ) \delta /\left({k + M} \right) represents the influence of the media reports, where \delta > 1 and k \geqslant 0 . \delta /k means the highest availability of media influence effect. \delta represents the rate of contact between the ignorant and the media reports and k is the effect of media reports. \gamma represents calm down rate of the spreader. \varepsilon and \omega represent the forgetting rates.

    (ⅲ) We define that the immigrate rate is \Lambda and the emigrate rate is \mu in our study. Furthermore, \eta represents the influence of rumors under the rumor control measures. {m_0} represents the initial value of the media reports. \rho represents the influence of rumor propagation on media reports. The parameters of the model are assumed to be constant and positive.

    The mathematical dynamics of rumor propagation under the Caputo sense is derived based on the propagation rules in Figure 1 as follows:

    \begin{array}{l} {D^\alpha }I\left( t \right) = \Lambda - \frac{{\beta I\left( t \right)S\left( t \right)}}{{1 + bS\left( t \right)}} - \frac{{\delta I\left( t \right)M\left( t \right)}}{{k + M\left( t \right)}} - \eta I\left( t \right) - \mu I\left( t \right), \\ {D^\alpha }S\left( t \right) = \frac{{\beta I\left( t \right)S\left( t \right)}}{{1 + bS\left( t \right)}} + \frac{{\delta I\left( t \right)M\left( t \right)}}{{k + M\left( t \right)}} - \gamma S\left( t \right) - \varepsilon S\left( t \right) - \mu S\left( t \right), \\ {D^\alpha }C\left( t \right) = \gamma S\left( t \right) - \omega C\left( t \right) - \mu C\left( t \right), \\ {D^\alpha }R\left( t \right) = \eta I\left( t \right) + \varepsilon S\left( t \right) + \omega C\left( t \right) - \mu R\left( t \right), \\ {D^\alpha }M\left( t \right) = {m_0} + \rho S\left( t \right) - \xi M\left( t \right), \\ \end{array} (3.1)

    where the initial conditions are denoted by \left({I\left(0 \right), S\left(0 \right), C\left(0 \right), R\left(0 \right), M\left(0 \right)} \right) = \left({{I_0}, {S_0}, {C_0}, {R_0}, {M_0}} \right) .

    Figure 1.  Flow diagram of rumor propagation.

    In this model, the non-negative initial conditions of Eq (3.1) are obtained as

    I\left( 0 \right) = {I_0} \geqslant 0, {\text{ }}S\left( 0 \right) = {S_0} \geqslant 0, {\text{ }}C\left( 0 \right) = {C_0} \geqslant 0, {\text{ }}R\left( 0 \right) = {R_0} \geqslant 0, {\text{ }}M\left( 0 \right) = {M_0} \geqslant 0. (3.2)

    Theorem 3.1. The solution of the Caputo sense Eq (3.1) subject to the conditions (3.2) is non-negative within the closed set

    \Omega = \left\{ {\left( {I, S, C, R, M} \right) \in R_ + ^4:N\left( t \right) = I\left( t \right) + S\left( t \right) + C\left( t \right)} \right. \left. { + R\left( t \right) \leqslant \frac{\Lambda }{\mu }, M\left( t \right) \leqslant \frac{{{m_0}}}{\xi } + \frac{{\rho \Lambda }}{{\mu \xi }}} \right\}.

    Proof. For Caputo fractional derivatives (3.1), one obtains

    \left\{ \begin{array}{l} {\left. {{D^\alpha }I\left( t \right)} \right|_{I\left( t \right) = 0}} = \Lambda \geqslant 0, \\ {\left. {{D^\alpha }S\left( t \right)} \right|_{S\left( t \right) = 0}} = \frac{{\delta I\left( t \right)M\left( t \right)}}{{k + M\left( t \right)}} \geqslant 0, \\ {\left. {{D^\alpha }S\left( t \right)} \right|_{S\left( t \right) = 0}} = \gamma S\left( t \right) \geqslant 0, \\ {\left. {{D^\alpha }R\left( t \right)} \right|_{R\left( t \right) = 0}} = \eta I\left( t \right) + \varepsilon S\left( t \right) + \omega C\left( t \right) \geqslant 0, \\ {\left. {{D^\alpha }M\left( t \right)} \right|_{M\left( t \right) = 0}} = {m_0} + \rho S\left( t \right) \geqslant 0. \\ \end{array} \right. (3.3)

    Based on the Laplace transform, it can be demonstrated that I, S, C, R, M \geqslant 0, {\text{ }}t \in \left[{{t_0}, T} \right). Additionally, the total population is determined as N\left(t \right) = I\left(t \right) + S\left(t \right) + C\left(t \right) + R\left(t \right) . Thus, adding the first four equations of Eq (3.1), we obtain

    \begin{array}{l} {}_0^CD_t^\alpha N\left( t \right) = \Lambda - \mu \left( {I\left( t \right) + S\left( t \right) + C\left( t \right) + R\left( t \right)} \right) \\ \;\;\;\;\;\;\;\;\;\;\;\;\;\; = \Lambda - \mu N\left( t \right). \\ \end{array}

    By employing the Laplace transform, we obtain N\left(t \right) \leqslant N\left(0 \right){E_\alpha }\left({ - \mu {t^\alpha }} \right) + \frac{\Lambda }{\mu }\left({1 - {E_\alpha }\left({ - \mu {t^\alpha }} \right)} \right). {E_\alpha }\left({ - \mu {t^\alpha }} \right) is the Mittag-Leffler function of parameter \alpha . Since 0 \leqslant {E_\alpha }\left({ - \mu {t^\alpha }} \right) \leqslant 1 , we obtain N\left(t \right) \leqslant \frac{\Lambda }{\mu } + N\left(0 \right). Subsequently, by applying the fifth equation of Eq (3.1), we obtain {D^\alpha }M\left(t \right) \leqslant {m_0} + \frac{{\rho \Lambda }}{\mu } - \xi M\left(t \right), and then, 0 \leqslant M\left(t \right) \leqslant \frac{{{m_0}}}{\xi } + \frac{{\rho \Lambda }}{{\mu \xi }} - M\left(0 \right), where M\left(0 \right) denotes the initial value of the message of media reports. Thus, \mathop {\sup }\limits_{t \to \infty } M\left(t \right) \leqslant \frac{{{m_0}}}{\xi } + \frac{\Lambda }{{\mu \xi }} . This implies that N\left(t \right) is bounded, thus I\left(t \right), S\left(t \right), C\left(t \right), R\left(t \right) and M\left(t \right) are bounded in

    \Omega = \left\{ {\left( {I, S, C, R, M} \right) \in R_ + ^4:N\left( t \right) = I\left( t \right) + S\left( t \right) + C\left( t \right) + R\left( t \right) \leqslant \frac{\Lambda }{\mu }, M\left( t \right) \leqslant \frac{{{m_0}}}{\xi } + \frac{{\rho \Lambda }}{{\mu \xi }}} \right\}.

    So, all solutions of Eq (3.1) are non-negative. Now, we will show that equation has a unique solution.

    Theorem 3.2. For any given initial condition X\left(0 \right) = \left({{I_0}, {H_0}, {S_0}, {R_0}, {M_0}} \right), {\text{ }}\forall t \geqslant 0 , Eq (3.1) has a unique solution X\left(t \right) \in \Omega .

    Proof. The approach employed in [40] is adopted to validate the existence and uniqueness.

    Let H\left(X \right) = \left({{H_1}\left(X \right), {H_2}\left(X \right), {H_3}\left(X \right), {H_4}\left(X \right), {H_5}\left(X \right)} \right) as follows:

    \begin{array}{l} {H_1}\left( X \right) = \Lambda - \frac{{\beta I\left( t \right)S\left( t \right)}}{{1 + bS\left( t \right)}} - \frac{{\delta I\left( t \right)M\left( t \right)}}{{k + M\left( t \right)}} - \eta I\left( t \right) - \mu I\left( t \right), \\ {H_2}\left( X \right) = \frac{{\beta I\left( t \right)S\left( t \right)}}{{1 + bS\left( t \right)}} + \frac{{\delta I\left( t \right)M\left( t \right)}}{{k + M\left( t \right)}} - \gamma S\left( t \right) - \varepsilon S\left( t \right) - \mu S\left( t \right), \\ {H_3}\left( X \right) = \gamma S\left( t \right) - \omega C\left( t \right) - \mu C\left( t \right), \\ {H_4}\left( X \right) = \eta I\left( t \right) + \varepsilon S\left( t \right) + \omega C\left( t \right) - \mu R\left( t \right), \\ {H_5}\left( X \right) = {m_0} + \rho S\left( t \right) - \xi M\left( t \right). \\ \end{array}

    Furthermore, for any given X, {X_1} \in \Omega , a function can be constructed as follows:

    \begin{array}{l} \left\| {H\left( X \right) - H\left( Y \right)} \right\| = \left| {{H_1}\left( X \right) - {H_1}\left( {{X_1}} \right)} \right| + \left| {{H_2}\left( X \right) - {H_2}\left( {{X_1}} \right)} \right| + \left| {{H_3}\left( X \right) - {H_3}\left( {{X_1}} \right)} \right| \\ \;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;{\text{ }} + \left| {{H_4}\left( X \right) - {H_4}\left( {{X_1}} \right)} \right| + \left| {{H_5}\left( X \right) - {H_5}\left( {{X_1}} \right)} \right| \\ \;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\; = \left| {\Lambda - \frac{{\beta IS}}{{1 + bS}} - \frac{{\delta IM}}{{k + M}} - \eta I - \mu I - \Lambda + \frac{{\beta {I_1}{S_1}}}{{1 + b{S_1}}} + \frac{{\delta {I_1}{M_1}}}{{k + {M_1}}} + \eta {I_1} + \mu {I_1}} \right| \\ \;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;{\text{ }} + \left| {\frac{{\beta IS}}{{1 + bS}} + \frac{{\delta IM}}{{k + M}} - \gamma S - \varepsilon S - \mu S - \frac{{\beta {I_1}{S_1}}}{{1 + b{S_1}}} - \frac{{\delta {I_1}{M_1}}}{{k + {M_1}}} + \gamma {S_1} + \varepsilon {S_1} + \mu {S_1}} \right| \\ \;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;{\text{ }} + \left| {\gamma S - \omega C - \mu C - \gamma {S_1} + \omega {C_1} + \mu {C_1}} \right| \\ \;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;{\text{ }}\; + \left| {\eta I + \varepsilon S + \omega C - \mu R - \eta {I_1} - \varepsilon {S_1} - \omega {C_1} + \mu {R_1}} \right| \\ \;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;{\text{ }} + \left| {{m_0} + \rho S - \xi M - {m_0} - \rho {S_1} + \xi {M_1}} \right| \\ \;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\; \leqslant {l_1}\left| {I - {I_1}} \right| + {l_2}\left| {S - {S_1}} \right| + {l_3}\left| {C - {C_1}} \right| + {l_4}\left| {R - {R_1}} \right| + {l_5}\left| {M - {M_1}} \right| \\ \;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\; \leqslant L\left\| {\left( {I, S, C, R, M} \right) - \left( {{I_1}, {S_1}, {C_1}, {R_1}, {M_1}} \right)} \right\| \\ \;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\; \leqslant L\left\| {X - {X_1}} \right\|, \\ \end{array}

    where

    \begin{array}{c}{l_1} = \frac{{2\beta }}{b} + 2\delta + 2\eta + \mu , {\text{ }}{l_2} = \frac{{2\beta \Lambda }}{\mu } + 2\gamma + 2\varepsilon + \rho + \mu , {\text{ }}{l_3} = 2\omega + \mu , {\text{ }}{l_4} = \mu , {\text{ }}{l_5} = \frac{{2\delta \Lambda }}{{k\mu }} + \xi ,\\ L = \max \left( {{l_1}, {l_2}, {l_3}, {l_4}, {l_5}} \right). \end{array}

    Therefore, H\left(X \right) satisfies the Lipschitz condition with any initial condition X\left(0 \right) = \left({{I_0}, {H_0}, {S_0}, {R_0}, {M_0}} \right), {\text{ }}\forall t \geqslant 0 . Equations (3.1) always have a unique solution.

    In order to explore the rumor-free and rumor-permanence equilibrium points of the fractional order Eq (3.1), we solve the following equations:

    \begin{array}{l} \Lambda - \frac{{\beta I\left( t \right)S\left( t \right)}}{{1 + bS\left( t \right)}} - \frac{{\delta I\left( t \right)M\left( t \right)}}{{k + M\left( t \right)}} - \eta I\left( t \right) - \mu I\left( t \right) = 0, \\ \frac{{\beta I\left( t \right)S\left( t \right)}}{{1 + bS\left( t \right)}} + \frac{{\delta I\left( t \right)M\left( t \right)}}{{k + M\left( t \right)}} - \gamma S\left( t \right) - \varepsilon S\left( t \right) - \mu S\left( t \right) = 0, \\ \gamma S\left( t \right) - \omega C\left( t \right) - \mu C\left( t \right) = 0, \\ \eta I\left( t \right) + \varepsilon S\left( t \right) + \omega C\left( t \right) - \mu R\left( t \right) = 0, \\ {m_0} + \rho S\left( t \right) - \xi M\left( t \right) = 0. \\ \end{array}

    By denoting the aforementioned Eq (3.1) with {S_0} = 0 , the rumor-free equilibrium point can be calculated as: {E_0} = \left({\frac{\Lambda }{{\eta + \mu }}, 0, 0, 0, \frac{{{m_0}}}{\xi } + \frac{{\rho \Lambda }}{{\xi \left({\eta + \mu } \right)}}} \right). By employing the next generation approach [41], the basic reproduction number {R_0} is calculated. If Q = {\left({S, C, M} \right)^T} , the original equation can be reformulated as

    \frac{{dQ}}{{dt}} = \mathcal{F}\left( x \right) - \mathcal{V}\left( x \right),

    where

    \mathcal{F}\left( x \right) = \left( \begin{array}{l} \frac{{\beta I\left( t \right)S\left( t \right)}}{{1 + bS\left( t \right)}} + \frac{{\delta I\left( t \right)M\left( t \right)}}{{k + M\left( t \right)}} \\ \;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;0 \\ \;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;0 \\ \end{array} \right), {\text{ }}\mathcal{V}\left( x \right) = \left( \begin{array}{l} \;\;\;\;\left( {\gamma + \varepsilon + \mu } \right)S\left( t \right) \\ \left( {\omega + \mu } \right)C\left( t \right) - \gamma S\left( t \right) \\ \;{m_0} + \xi M\left( t \right) - \rho S\left( t \right) \\ \end{array} \right).

    The Jacobian matrices of \mathcal{F}\left(x \right) and \mathcal{V}\left(x \right) at the rumor-free equilibrium {E_0} , respectively, are

    {F_0} = \left( \begin{array}{l} 0\;\;\;\;\beta {I_0}\;\;\;\;0\;\;\;\;\;\;0\;\;\;\;\;\frac{{\delta {I_0}}}{k} \\ 0\;\;\;\;\;0\;\;\;\;\;\;\;0\;\;\;\;\;\;0\;\;\;\;\;\;\;0 \\ 0\;\;\;\;\;0\;\;\;\;\;\;\;0\;\;\;\;\;\;0\;\;\;\;\;\;\;0 \\ \end{array} \right), {\text{ }}{V_0} = \left( \begin{array}{l} 0\;\;\gamma + \varepsilon + \mu \;\;\;\;\;\;0\;\;\;\;\;\;0\;\;\;\;0 \\ 0\;\;\;\;\; - \gamma \;\;\;\;\;\;\;\;\omega + \mu \;\;\;0\;\;\;\;0 \\ 0\;\;\;\; - \rho \;\;\;\;\;\;\;\;\;\;\;0\;\;\;\;\;\;\;0\;\;\;\;\xi \; \\ \end{array} \right).

    The next-generation matrix for Eq (3.1) is presented as follows:

    {F_0}V_0^{ - 1} = \left( \begin{array}{l} \;\;\;\;\;\;\;\;\;\;\;\;0\;\;\;\;\;\;\;\;\;\;\;\;\;\;\; 0 \;\;\;\;\;\; 0 \\ \frac{{\beta {I_0}}}{{\gamma + \varepsilon + \mu }} + \frac{{\rho \delta {I_0}}}{{k\xi \left( {\gamma + \varepsilon + \mu } \right)}}\;\;\;\;0\;\;\;\;\frac{{\delta {I_0}}}{{k\xi }} \\ \;\;\;\;\;\;\;\;\;\;\;\;\;0\;\;\;\;\;\;\;\;\; \;\;\;\;\;0\;\;\;\;\;\;0 \\ \end{array} \right).

    Subsequently, the basic reproduction number of Eq (3.1) is determined:

    {R_0} = \rho \left( {{F_0}V_0^{ - 1}} \right) = \frac{{\beta {I_0}}}{{\gamma + \varepsilon + \mu }} + \frac{{\rho \delta {I_0}}}{{k\xi \left( {\gamma + \varepsilon + \mu } \right)}} = \frac{{\Lambda \left( {k\xi \beta + \rho \delta } \right)}}{{k\xi \left( {\eta + \mu } \right)\left( {\gamma + \varepsilon + \mu } \right)}}.

    The following theorem is utilized to analyze the local and global stability of the two equilibrium points.

    Theorem 4.1. Equation (3.1) exhibit local stability of the rumor-free equilibrium {E_0} if {R_0} < 1 , and if {R_0} > 1 it is unstable.

    Proof. The stability of the equilibrium point is investigated by examining the Jacobian matrix of Eq (3.1).

    J\left( {{E_0}} \right) = \left( \begin{array}{l} - \eta - \mu \;\;\;\;\;\;\;\;\;\;\;\; - \beta {I_0}\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;0\;\;\;\;\;\;\;\;\;\;\;0\;\;\;\;\; - \frac{{\delta {I_0}}}{k} \\ \;\;\;\;0\;\;\;\;\;\;\beta {I_0} - \left( {\gamma + \varepsilon + \mu } \right)\;\;\;\;\;\;\;\;\;\;0\;\;\;\;\;\;\;\;\;\;\;0\;\;\;\;\;\;\frac{{\delta {I_0}}}{k} \\ \;\;\;\;0\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\gamma \;\;\;\;\;\;\;\;\;\;\;\;\; - \omega - \mu \;\;\;\;\;\;0\;\;\;\;\;\;\;\;0 \\ \;\;\;\;\eta \;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\varepsilon \;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\omega \;\;\;\;\;\;\;\;\; - \mu \;\;\;\;\;\;0\;\;\;\; \\ \;\;\;\;0\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\rho \;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;0\;\;\;\;\;\;\;\;\;\;\;\;0\;\;\;\;\;\; - \xi \\ \end{array} \right).

    The characteristic polynomial of J\left({{E_0}} \right) is derived as follows:

    Det\left( {\lambda I - J\left( {{E_0}} \right)} \right) = \left( {\lambda + \mu } \right)\left( {\lambda + \eta + \mu } \right)\left( {\lambda + \omega + \mu } \right)\left[ {\left( {\lambda - \beta {I_0} + \left( {\gamma + \varepsilon + \mu } \right)} \right)\left( {\lambda + \xi } \right) - \frac{{\rho \delta {I_0}}}{k}} \right]. (4.1)

    If all roots of Det\left({\lambda I - J\left({{E_0}} \right)} \right) possess negative real parts, the rumor-free equilibrium {E_0} can be considered asymptotically stable.

    When Det\left({\lambda I - J\left({{E_0}} \right)} \right) = 0 , we obtain

    {\lambda _1} = - \mu , {\text{ }}{\lambda _2} = - \eta - \mu , {\text{ }}{\lambda _3} = - \omega - \mu ,

    and

    {\lambda ^2} + \left( {\gamma + \varepsilon + \mu + \xi - \beta {I_0}} \right)\lambda + \xi \left( {\gamma + \varepsilon + \mu } \right) - \xi \beta {I_0} - \frac{{\rho \delta {I_0}}}{k} = 0. (4.2)

    It is obvious that {\lambda _1}, {\text{ }}{\lambda _2} and {\lambda _3} are negatives roots and thus, the real component of each of Eq (4.2) roots plays a crucial role.

    According to Eq (4.1), it defines

    {b_0} = 1, {\text{ }}{b_1} = \gamma + \varepsilon + \mu + \xi - \beta {I_0}, {\text{ }}{b_2} = \xi \left( {\gamma + \varepsilon + \mu } \right) - \xi \beta {I_0} - \frac{{\rho \delta {I_0}}}{k}.

    Evidently, we obtain

    {b_0}{\lambda ^2} + {b_1}\lambda + {b_2} = 0. (4.3)

    When {R_0} < 1 , we acquire

    \xi \left( {\gamma + \varepsilon + \mu } \right) - \xi \beta {I_0} - \frac{{\rho \delta {I_0}}}{k} > 0. (4.4)

    Thus, {b_2} > 0 is deducted from Eq (4.2) by adding \frac{{\rho \delta {I_0}}}{k} to Eq (4.4), and we obtain

    \xi \left( {\gamma + \varepsilon + \mu } \right) - \xi \beta {I_0} > \frac{{\rho \delta {I_0}}}{k} > 0. (4.5)

    From Eq (4.5), we obtain

    \gamma + \varepsilon + \mu - \beta {I_0} > \frac{{\rho \delta {I_0}}}{{k\xi }} > 0. (4.6)

    By adding the parameter \xi to Eq (4.6), we obtain

    \gamma + \varepsilon + \mu + \xi - \beta {I_0} > \xi > 0. (4.7)

    Therefore, {b_1} > 0 is deduced based on Eq (4.7). After evaluation of different coefficients of Eq (4.2), {b_0} > 0, {\text{ }}{b_1} > 0, {\text{ }}{b_2} > 0 and {b_1}{b_2} > 0 are obtained. According to the Routh-Hurwitz conditions [42], all eigenvalues of J\left({{E_0}} \right) have negative real part. The condition \left| {\arg \left({{\lambda _i}} \right)} \right| \geqslant \frac{{\alpha \pi }}{2} is satisfied and the rumor-free equilibrium {E_0} is locally asymptotically stable if {R_0} < 1 , and unstable if {R_0} > 1 .

    Theorem 4.2. The rumor-free equilibrium {E_0} of Eq (3.1) is globally asymptotically stable if {R_0} < 1 .

    Proof. Let V\left({I, S, C, R, M} \right) , with positive constants {c_1} and {c_2} , serve as a Lyapunov function

    V\left( {I, S, C, R, M} \right) = \left( {I - {I_0} - {I_0}\ln \frac{I}{{{I_0}}}} \right) + \left( {C - {C_0} - {C_0}\ln \frac{C}{{{C_0}}}} \right)\; + \left( {R - {R_0} - {R_0}\ln \frac{R}{{{R_0}}}} \right) + {c_1}S + {c_2}M.

    By differentiating the Lyapunov function with respect to time, we obtain

    {D^\alpha }V = \left( {1 - \frac{{{I_0}}}{I}} \right){D^\alpha }I\left( t \right) + \left( {1 - \frac{{{C_0}}}{C}} \right){D^\alpha }C\left( t \right) + \left( {1 - \frac{{{R_0}}}{R}} \right){D^\alpha }R\left( t \right)\; + {c_1}{D^\alpha }S\left( t \right) + {c_2}{D^\alpha }M\left( t \right), (4.8)

    where {c_1} \geqslant 0, {\text{ }}{c_2} \geqslant 0. Substituting {D^\alpha }I\left(t \right), {D^\alpha }S\left(t \right), {D^\alpha }C\left(t \right), {D^\alpha }R\left(t \right) , and {D^\alpha }M\left(t \right) in Eq (3.1) into Eq (4.8) yields

    \begin{array}{l} {D^\alpha }V = \left( {1 - \frac{{{I_0}}}{I}} \right)\left[ {\Lambda - \frac{{\beta I\left( t \right)S\left( t \right)}}{{1 + bS\left( t \right)}} - \frac{{\delta I\left( t \right)M\left( t \right)}}{{k + M\left( t \right)}} - \eta I\left( t \right) - \mu I\left( t \right)} \right] \\ \;\;\;\;\;\;\;\;\;\;\;\; + \left( {1 - \frac{{{C_0}}}{C}} \right)\left[ {\gamma S\left( t \right) - \omega C\left( t \right) - \mu C\left( t \right)} \right] \\ \;\;\;\;\;\;\;\;\;\;\;\; + \left( {1 - \frac{{{R_0}}}{R}} \right)\left[ {\eta I\left( t \right) + \varepsilon S\left( t \right) + \omega C\left( t \right) - \mu R\left( t \right)} \right] \\ \;\;\;\;\;\;\;\;\;\;\;\; + {c_1}\left[ {\frac{{\beta I\left( t \right)S\left( t \right)}}{{1 + bS\left( t \right)}} + \frac{{\delta I\left( t \right)M\left( t \right)}}{{k + M\left( t \right)}} - \gamma S\left( t \right) - \varepsilon S\left( t \right) - \mu S\left( t \right)} \right] \\ \;\;\;\;\;\;\;\;\;\;\;\; + {c_2}\left[ {\rho S\left( t \right) - \xi M\left( t \right)} \right], \\ \end{array}
    {D^\alpha }V \leqslant \left[ {\frac{{{c_1}\delta I\left( t \right)}}{{k + M\left( t \right)}} - {c_2}\xi } \right]M\left( t \right) + \left[ {\frac{{{c_1}\beta I\left( t \right)}}{{1 + bS\left( t \right)}} - {c_1}\left( {\gamma + \varepsilon + \mu } \right) + {c_2}\rho } \right]S\left( t \right). (4.9)

    From Eq (4.9), we obtain

    {d_1} = \frac{{{c_1}\delta I\left( t \right)}}{{k + M\left( t \right)}} - {c_2}\xi , {\text{ }}{d_2} = \frac{{{c_1}\beta I\left( t \right)}}{{1 + bS\left( t \right)}} - {c_1}\left( {\gamma + \varepsilon + \mu } \right) + {c_2}.

    Equating the coefficient {d_1} to zero, we obtain

    \frac{{{c_1}\delta I\left( t \right)}}{{k + M\left( t \right)}} = {c_2}\xi .

    At the rumor-free equilibrium {E_0} , I \leqslant {I^0} and M \leqslant {M^0} = 0 , we obtain \frac{{{c_1}\delta {I_0}}}{k} = {c_2}\xi . By considering {c_1} = \xi, {\text{ }}{c_2} = \frac{{\delta {I_0}}}{k}, and substituting {c_1} and {c_2} into Eq (4.9), the obtained result is as follows:

    \begin{array}{l} {c_2} = \frac{{\xi \beta {I_0}}}{{1 + bS\left( t \right)}} + \frac{{\rho \delta {I_0}}}{k} - \xi \left( {\gamma + \varepsilon + \mu } \right) \\ \;\;\;\; \leqslant \xi \beta {I_0} + \frac{{\rho \delta {I_0}}}{k} - \xi \left( {\gamma + \varepsilon + \mu } \right){\text{ }}\left( {for{\text{ }}1 + bS\left( t \right) \geqslant 1} \right) \\ \;\;\;\; = \xi \left( {\gamma + \varepsilon + \mu } \right)\left( {{R_0} - 1} \right). \\ \end{array}

    If {R_0} \leqslant 1 , then {D^\alpha }V\left(t \right) \leqslant 0 . Specially, {D^\alpha }V\left(t \right) = 0 if and only if S\left(t \right) = 0 . Therefore, the largest compact invariant set in \left\{ {V\left({I, S, C, R, M} \right) \in \Omega :{D^\alpha }V\left(t \right) \leqslant 0} \right\} is the singleton set \left\{ {{E_0}} \right\} . According to LaSalle's invariance principle [43], {E_0} is globally asymptotically stable in \Omega if {R_0} \leqslant 0 .

    Theorem 4.3. If {R_0} > 1 and {l_1}{l_2} - {l_3} > 0, {\text{ }}{l_1}{l_2}{l_3} - l_1^2{l_3} - l_3^2 > 0 , Eq (3.1) exhibits local stability of rumor-permanence equilibrium {E^*} .

    Proof. Equation (3.1) at the rumor-permanence equilibrium {E^*} can be further simplified as follows:

    \begin{array}{l} \Lambda - \frac{{\beta {I^*}{S^*}}}{{1 + b{S^*}}} - \frac{{\delta {I^*}{M^*}}}{{k + {M^*}}} - \eta {I^*} - \mu {I^*} = 0, \\ \frac{{\beta {I^*}{S^*}}}{{1 + b{S^*}}} + \frac{{\delta {I^*}{M^*}}}{{k + {M^*}}} - \gamma {S^*} - \varepsilon {S^*} - \mu {S^*} = 0, \\ \gamma {S^*} - \omega {C^*} - \mu {C^*} = 0, \\ {m_0} + \rho {S^*} - \xi {M^*} = 0. \\ \end{array}

    The Jacobian matrix evaluated at the rumor-permanence equilibrium {E^*} persistence is presented in [16] as

    J\left({E}^{*}\right) = \left(\begin{array}{l}{J}_{11}^{*}{J}_{12}^{*}0{J}_{14}^{*}\\ {J}_{21}^{*}{J}_{22}^{*}0{J}_{24}^{*}\\ 0{J}_{32}^{*}{J}_{33}^{*}0\\ 0{J}_{42}^{*}0{J}_{44}^{*}\end{array}\right),

    where

    \begin{array}{l} J_{11}^* = - \frac{{\beta {S^*}}}{{1 + b{S^*}}} - \frac{{\delta {M^*}}}{{k + {M^*}}} - \eta - \mu , {\text{ }}J_{12}^* = - \frac{{\beta {I^*}}}{{{{\left( {1 + b{S^*}} \right)}^2}}}, {\text{ }}J_{14}^* = - \frac{{k\delta {I^*}}}{{{{\left( {k + {M^*}} \right)}^2}}}, \\ J_{21}^* = \frac{{\beta {S^*}}}{{1 + b{S^*}}} + \frac{{\delta {M^*}}}{{k + {M^*}}}, {\text{ }}J_{22}^* = \frac{{\beta {I^*}}}{{{{\left( {1 + b{S^*}} \right)}^2}}} - \gamma - \varepsilon - \mu , {\text{ }}J_{24}^* = \frac{{k\delta {I^*}}}{{{{\left( {k + {M^*}} \right)}^2}}}, \\ J_{32}^* = \gamma , {\text{ }}J_{33}^* = - \omega , {\text{ }}J_{42}^* = \rho , {\text{ }}J_{44}^* = - \xi . \\ \end{array}

    The corresponding characteristic equation can be derived as

    {\lambda ^4} + {l_1}{\lambda ^3} + {l_2}{\lambda ^2} + {l_3}\lambda + {l_4} = 0, (4.10)

    where

    \begin{array}{l} {l_1} = - J_{11}^* - J_{22}^* - J_{33}^* - J_{44}^*, \\ {l_2} = J_{11}^*J_{22}^* - J_{12}^*J_{21}^* + J_{11}^*J_{33}^* + J_{11}^*J_{44}^* + J_{22}^*J_{33}^* + J_{22}^*J_{44}^* - J_{24}^*J_{42}^* + J_{33}^*J_{44}^*, \\ {l_3} = - J_{11}^*J_{22}^*J_{33}^* + J_{12}^*J_{21}^*J_{33}^* - J_{11}^*J_{24}^*J_{42}^* + J_{12}^*J_{21}^*J_{44}^* - J_{14}^*J_{21}^*J_{42}^* \\ \;\;\;\;\;\; - J_{11}^*J_{33}^*J_{44}^* - J_{22}^*J_{33}^*J_{44}^* + J_{24}^*J_{33}^*J_{42}^*, \\ {l_4} = J_{11}^*J_{22}^*J_{33}^*J_{44}^* - J_{11}^*J_{24}^*J_{33}^*J_{42}^* - J_{12}^*J_{21}^*J_{33}^*J_{44}^* + J_{14}^*J_{21}^*J_{33}^*J_{42}^*. \\ \end{array}

    For {R_0} > 1 , J_{ii}^*{\text{ }}\left({i = 1, 2, 3, 4} \right), J_{12}^* and J_{14}^* are all negative, while J_{21}^*, J_{24}^*, J_{32}^* and J_{42}^* are positive. Subsequently, we obtain

    \begin{array}{l} {\Delta _1} = {l_1} = - J_{11}^* - J_{22}^* - J_{33}^* - J_{44}^* > 0, \\ {\Delta _2} = \left| \begin{array}{l} {l_1}\;\;\;\;1 \\ {l_3}\;\;\;\;{l_2} \\ \end{array} \right| = {l_1}{l_2} - {l_3}, \\ {\Delta _3} = \left( \begin{array}{l} {l_1}\;\;\;\;1\;\;\;\;0 \\ {l_3}\;\;\;\;{l_2}\;\;\;{l_1} \\ 0\;\;\;\;{l_4}\;\;\;\;{l_3} \\ \end{array} \right) = {l_1}{l_2}{l_3} - l_1^2{l_3} - l_3^2. \\ \end{array}

    Obviously, it is {\Delta _1} > 0 . If {l_1}{l_2} - {l_3} > 0, {\text{ }}{l_1}{l_2}{l_3} - l_1^2{l_3} - l_3^2 > 0 , then {\Delta _2} > 0, {\text{ }}{\Delta _3} > 0 . According to the Routh-Hurwitz criterion [42], it can be inferred that all eigenvalues of the eigenequation in Eq (4.10) possess negative real parts.

    Therefore, the rumor-permanence equilibrium {E^*} is proven to be locally asymptotically stable and Theorem 4.3 is proven.

    The objective of our study is to identify the optimal control method and approach in order to minimize the overall cost associated with rumor spreading. The feasible intervention method aims to minimize the dissemination of rumors by individuals. The primary objective of fractional optimal control is to mitigate the relative density of individuals who spread rumors. Furthermore, the consideration of cost control measures is imperative in light of the current circumstances. Therefore, the optimal control strategy for Eq (5.1) is proposed, and its investigation can be facilitated through the application of Pontryagin's maximum principle [44].

    Equation (3.1), incorporating the control mechanism, is thus formulated as

    \begin{array}{l} {D^\alpha }I\left( t \right) = \Lambda - \frac{{\beta \left( {1 - {u_1}\left( t \right)} \right)I\left( t \right)S\left( t \right)}}{{1 + bS\left( t \right)}} - \frac{{\delta I\left( t \right)M\left( t \right)}}{{k + M\left( t \right)}} - \eta I\left( t \right) - \mu I\left( t \right), \\ {D^\alpha }S\left( t \right) = \frac{{\beta \left( {1 - {u_1}\left( t \right)} \right)I\left( t \right)S\left( t \right)}}{{1 + bS\left( t \right)}} + \frac{{\delta I\left( t \right)M\left( t \right)}}{{k + M\left( t \right)}} - \gamma S\left( t \right) - \varepsilon {u_2}\left( t \right)S\left( t \right) - \mu S\left( t \right), \\ {D^\alpha }C\left( t \right) = \gamma S\left( t \right) - \omega {u_3}\left( t \right)C\left( t \right) - \mu C\left( t \right), \\ {D^\alpha }R\left( t \right) = \eta I\left( t \right) + \varepsilon {u_2}\left( t \right)S\left( t \right) + \omega {u_3}\left( t \right)C\left( t \right) - \mu R\left( t \right), \\ {D^\alpha }M\left( t \right) = {m_0} + \rho S\left( t \right) - \xi M\left( t \right), \\ \end{array} (5.1)

    where the control strengths to the rumor propagation are represented by {u_1}\left(t \right), {\text{ }}{u_2}\left(t \right) and {u_3}\left(t \right) through education or government management. The objective function of the optimal control is formulated as

    Minimize \;J\left({u}_{1}, {u}_{2}, {u}_{3}\right) = {\displaystyle {\int }_{0}^{T}\left({A}_{1}S\left(t\right)+{A}_{2}C\left(t\right)+{K}_{1}\frac{{u}_{1}^{2}}{2}+{K}_{2}\frac{{u}_{2}^{2}}{2}+{K}_{3}\frac{{u}_{3}^{2}}{2}\right)dt}\text{,} (5.2)

    where {A_1} and {A_2} are the non-negative weights related to S\left(t \right) and C\left(t \right) in Eq (5.2). {K_1}, {\text{ }}{K_2}, and {K_3} are relative to cost of the measures.

    To determine the optimal interventions proposed in Eq (5.1),

    U\left\{ {\left. {\left( {{u_1}\left( t \right), {u_2}\left( t \right), {u_3}\left( t \right)} \right)} \right|0 \leqslant {u_1}\left( t \right) \leqslant u_1^{\max }, 0 \leqslant {u_2}\left( t \right) \leqslant u_2^{\max }, 0 \leqslant {u_3}\left( t \right) \leqslant u_3^{\max }, t \in \left( {0, T} \right]} \right\}, (5.3)

    where u_1^{\max } < 1, {\text{ }}u_2^{\max } < 1, and u_3^{\max } < 1 are the upper bounds of {u_1}\left(t \right), {\text{ }}{u_2}\left(t \right), and {u_3}\left(t \right) , respectively. Optimal controls u_1^*, u_2^*, and u_3^* satisfy J\left({u_1^*, u_2^*, u_2^*} \right) = \min \left\{ {J\left({{u_1}\left(t \right), {u_2}\left(t \right), {u_3}\left(t \right)} \right):} \right. \left. {\left({{u_1}\left(t \right), {u_2}\left(t \right), {u_3}\left(t \right)} \right) \in U} \right\} .

    We derive the optimal controls for minimizing the rumor propagation at minimal cost. Consequently, the following Lagrangian function is obtained:

    \begin{array}{l}{\rm H} = {A}_{1}S\left(t\right)+{A}_{2}C\left(t\right)+{K}_{1}\frac{{u}_{1}^{2}}{2}+{K}_{2}\frac{{u}_{2}^{2}}{2}+{K}_{3}\frac{{u}_{3}^{2}}{2}\\ \text{ }+{\lambda }_{1}\left(\Lambda -\frac{\beta \left(1-{u}_{1}\left(t\right)\right)I\left(t\right)S\left(t\right)}{1+bS\left(t\right)}-\frac{\delta I\left(t\right)M\left(t\right)}{k+M\left(t\right)}-\eta I\left(t\right)-\mu I\left(t\right)\right)\\ \text{ }+{\lambda }_{2}\left(\frac{\beta \left(1-{u}_{1}\left(t\right)\right)I\left(t\right)S\left(t\right)}{1+bS\left(t\right)}+\frac{\delta I\left(t\right)M\left(t\right)}{k+M\left(t\right)}-\left(\gamma +\varepsilon {u}_{2}\left(t\right)+\mu \right)S\left(t\right)\right)\\ \text{ }+{\lambda }_{3}\left(\gamma S\left(t\right)-\omega {u}_{3}\left(t\right)C\left(t\right)-\mu C\left(t\right)\right)\\ \text{ }+{\lambda }_{4}\left(\eta I\left(t\right)+\varepsilon {u}_{2}\left(t\right)S\left(t\right)+\omega {u}_{3}\left(t\right)C\left(t\right)-\mu R\left(t\right)\right)\\ \text{ }+{\lambda }_{5}\left({m}_{0}+\rho S\left(t\right)-\xi M\left(t\right)\right)\text{,}\end{array} (5.4)

    where {\lambda _1}, {\lambda _2}, {\lambda _3}, {\lambda _4}, {\lambda _5} are adjoint variable functions.

    Theorem 5.1. There exists an optimal control with the solution \left({I, S, C, R, M} \right) corresponding to the state equation in Eq (5.1) and the adjoint variables {\lambda _1}, {\lambda _2}, {\lambda _3}, {\lambda _4}, {\lambda _5} , we obtain

    \begin{array}{l} {D^\alpha }{\lambda _1} = - {\lambda _1}\left( {\frac{{\beta \left( {1 - {u_1}} \right)S}}{{1 + bS}} - \frac{{\delta M}}{{k + M}} + \eta + \mu } \right) - {\lambda _2}\rho \left( {\frac{{\beta \left( {1 - {u_1}} \right)S}}{{1 + bS}} - \frac{{\delta M}}{{k + M}}} \right) + {\lambda _4}\eta , \\ {D^\alpha }{\lambda _2} = \frac{{ - {\lambda _1}\beta \left( {1 - {u_1}} \right)I}}{{{{\left( {1 + bS} \right)}^2}}} + {\lambda _2}\left( {\frac{{\beta \left( {1 - {u_1}} \right)I}}{{{{\left( {1 + bS} \right)}^2}}} - \left( {\gamma + \varepsilon {u_2}\left( t \right) + \mu } \right)} \right) - {\lambda _3}\gamma - {\lambda _4}\varepsilon {u_2}\left( t \right) - \;{\lambda _5}\rho , \\ {D^\alpha }{\lambda _3} = {\lambda _3}\left( {\omega {u_3}\left( t \right) + \mu } \right) - {\lambda _4}\omega {u_3}\left( t \right), \\ {D^\alpha }{\lambda _4} = {\lambda _4}\mu , \\ {D^\alpha }\;{\lambda _5} = {\lambda _1}\frac{{k\delta I}}{{{{\left( {k + M} \right)}^2}}} - {\lambda _2}\frac{{k\delta I}}{{{{\left( {k + M} \right)}^2}}} + \;{\lambda _5}\xi M, \\ \end{array}

    with boundary conditions {\lambda _1}\left(t \right) = {\lambda _2}\left(t \right) = {\lambda _3}\left(t \right) = {\lambda _4}\left(t \right) = {\lambda _5}\left(t \right) = 0. According to the adjoint variables I\left(t \right), S\left(t \right), C\left(t \right), R\left(t \right) and M\left(t \right) , we differentiate the Hamiltonian {\rm H} in Eq (5.2) with respect to each variable.

    Proof. From the adjoint variables I\left(t \right), S\left(t \right), C\left(t \right), R\left(t \right) and M\left(t \right) , we differentiate the Hamiltonian {\rm H} in Eq (5.2) with respect to each variable:

    \begin{array}{l} {D^\alpha }{\lambda _1} = \frac{{ - \partial {\rm H}}}{{\partial I}} = {\lambda _1}\left( {\frac{{\beta \left( {1 - {u_1}} \right)S}}{{1 + bS}} - \frac{{\delta M}}{{k + M}} + \eta + \mu } \right) - {\lambda _2}\rho \left( {\frac{{\beta \left( {1 - {u_1}} \right)S}}{{1 + bS}} - \frac{{\delta M}}{{k + M}}} \right) + {\lambda _4}\eta , \\ {D^\alpha }{\lambda _2} = \frac{{ - \partial {\rm H}}}{{\partial S}} = \frac{{ - {\lambda _1}\beta \left( {1 - {u_1}} \right)I}}{{{{\left( {1 + bS} \right)}^2}}} + {\lambda _2}\left( {\frac{{\beta \left( {1 - {u_1}} \right)I}}{{{{\left( {1 + bS} \right)}^2}}} - \left( {\gamma + \varepsilon {u_2}\left( t \right) + \mu } \right)} \right) - {\lambda _3}\gamma - {\lambda _4}\varepsilon {u_2}\left( t \right) - \;{\lambda _5}\rho , \\ {D^\alpha }{\lambda _3} = \frac{{ - \partial {\rm H}}}{{\partial C}} = {\lambda _3}\left( {\omega {u_3}\left( t \right) + \mu } \right) - {\lambda _4}\omega {u_3}\left( t \right), \\ {D^\alpha }{\lambda _4} = \frac{{ - \partial {\rm H}}}{{\partial R}} = {\lambda _4}\mu , \\ {D^\alpha }\;{\lambda _5} = \frac{{ - \partial {\rm H}}}{{\partial M}} = {\lambda _1}\frac{{k\delta I}}{{{{\left( {k + M} \right)}^2}}} - {\lambda _2}\frac{{k\delta I}}{{{{\left( {k + M} \right)}^2}}} + \;{\lambda _5}\xi M. \\ \end{array}

    Moreover, the optimal control of the control variables set {u_1}, {u_2}, {u_3} is attained by partial differential equation as follows:

    \frac{{\partial {\rm H}\left( {I, S, C, R, M} \right)}}{{\partial {u_i}}} = 0, {\text{ }}i = 1, 2, 3.

    For u_1^*, {\text{ }}\frac{{\partial {\rm H}\left({I, S, C, R, M} \right)}}{{\partial {u_1}}} = 0,

    u_1^* = \frac{{\beta SI\left( {{\lambda _1} - {\lambda _2}} \right)}}{{{K_1}}}.

    For u_2^*, \frac{{\partial {\rm H}\left({I, S, C, R, M} \right)}}{{\partial {u_2}}} = 0,

    u_2^* = \frac{{\varepsilon S\left( { - {\lambda _2} + \;{\lambda _4}} \right)}}{{{K_2}}}.

    For u_3^*, \frac{{\partial {\rm H}\left({I, S, C, R, M} \right)}}{{\partial {u_3}}} = 0,

    u_3^* = \frac{{\omega C\left( { - {\lambda _3} + {\lambda _4}} \right)}}{{{K_3}}}.

    Therefore,

    \begin{array}{l} u_1^* = \max \left\{ {0, \min \left( {1, \frac{{\omega C\left( { - {\lambda _3} + {\lambda _4}} \right)}}{{{K_3}}}} \right)} \right\}, \\ u_2^* = \max \left\{ {0, \min \left( {1, \frac{{\varepsilon S\left( { - {\lambda _2} + \;{\lambda _4}} \right)}}{{{K_2}}}} \right)} \right\}, \\ u_3^* = \max \left\{ {0, \min \left( {1, \frac{{\omega C\left( { - {\lambda _3} + {\lambda _4}} \right)}}{{{K_3}}}} \right)} \right\}. \\ \end{array}

    The uniqueness of the optimal control has been established based on the boundedness and adjoint variables. The standard control with a bound is obtained as follows:

    {u}_{1}^{*} = \left\{ \begin{array}{l}0, \;\;\;\;\;\;\;\;\;if\;\;\;\frac{\beta SI\left({\lambda }_{1}-{\lambda }_{2}\right)}{{K}_{1}} < 0, \\ \left(0, 1\right),\;\;\;if\;\;\;\frac{\beta SI\left({\lambda }_{1}-{\lambda }_{2}\right)}{{K}_{1}}\le 1, \\ 1, \;\;\;\;\;\;\;\;\;if\;\;\;\frac{\beta SI\left({\lambda }_{1}-{\lambda }_{2}\right)}{{K}_{1}} > 1, \end{array} \right.
    {u}_{2}^{*} = \left\{ \begin{array}{l}0, \;\;\;\;\;\;\;\;\;if\;\;\;\frac{\varepsilon S\left(-{\lambda }_{2}+{\lambda }_{4}\right)}{{K}_{2}} < 0, \\ \left(0, 1\right),\;\;\;if\;\;\;\frac{\varepsilon S\left(-{\lambda }_{2}+{\lambda }_{4}\right)}{{K}_{2}}\le 1, \\ 1, \;\;\;\;\;\;\;\;\;if\;\;\;\frac{\varepsilon S\left(-{\lambda }_{2}+{\lambda }_{4}\right)}{{K}_{2}} > 1, \end{array} \right.
    {u}_{3}^{*} = \left\{ \begin{array}{l}0, \;\;\;\;\;\;\;\;\;if\;\;\;\frac{\omega C\left(-{\lambda }_{3}+{\lambda }_{4}\right)}{{K}_{3}} < 0, \\ \left(0, 1\right),\;\;\;if\;\;\;\frac{\omega C\left(-{\lambda }_{3}+{\lambda }_{4}\right)}{{K}_{3}}\le 1, \\ 1, \;\;\;\;\;\;\;\;\;if\;\;\;\frac{\omega C\left(-{\lambda }_{3}+{\lambda }_{4}\right)}{{K}_{3}} > 1.\end{array} \right.

    The sensitivity of the parameters in the rumor propagation model is investigated in this section. The following conditions are satisfied:

    \begin{array}{l} {A_\beta } = \frac{\beta }{{{R_0}}}\frac{{\partial {R_0}}}{{\partial \beta }} = \frac{{k\xi \beta }}{{k\xi \beta + \rho \delta }} > 0, \\ {A_\rho } = \frac{\rho }{{{R_0}}}\frac{{\partial {R_0}}}{{\partial \rho }} = \frac{{\rho \delta }}{{k\xi \beta + \rho \delta }} > 0, \\ {A_\delta } = \frac{\delta }{{{R_0}}}\frac{{\partial {R_0}}}{{\partial \delta }} = \frac{{\rho \delta }}{{k\xi \beta + \rho \delta }} > 0, \\ {A_\eta } = \frac{\eta }{{{R_0}}}\frac{{\partial {R_0}}}{{\partial \eta }} = - \frac{\eta }{{\left( {\eta + \mu } \right)}} < 0. \\ \end{array}

    In conclusion, \beta , \rho , and \delta are sensitive, and \eta , analogous to \gamma and \varepsilon are not sensitive. Sensitivity analysis shows that it is possible to reduce the rumor contact rate \beta and parameters \rho and \delta , thus lowering the value of {R_0} .

    In this section, we present the fractional Euler method [45], which is employed in solving fractional-order differential equations and obtaining a discretized equation, as described below:

    \begin{array}{l} I\left( {{t_p}} \right) = I\left( {{t_0}} \right) + \frac{{{h^\alpha }}}{{\Gamma \left( {\alpha + 1} \right)}}\sum\limits_{j = 0}^{p - 1} {{c_{j, p}}\left[ {\Lambda - \frac{{\beta I\left( {{t_j}} \right)S\left( {{t_j}} \right)}}{{1 + bS\left( {{t_j}} \right)}} - \frac{{\delta I\left( {{t_j}} \right)M\left( {{t_j}} \right)}}{{k + M\left( {{t_j}} \right)}} - \left( {\eta + \mu } \right)I\left( {{t_j}} \right)} \right]} , \\ S\left( {{t_p}} \right) = S\left( {{t_0}} \right) + \frac{{{h^\alpha }}}{{\Gamma \left( {\alpha + 1} \right)}}\sum\limits_{j = 0}^{p - 1} {{c_{j, p}}\left[ {\frac{{\beta I\left( {{t_j}} \right)S\left( {{t_j}} \right)}}{{1 + bS\left( {{t_j}} \right)}} + \frac{{\delta I\left( {{t_j}} \right)M\left( {{t_j}} \right)}}{{k + M\left( {{t_j}} \right)}} - \left( {\gamma + \varepsilon + \mu } \right)S\left( {{t_j}} \right)} \right]} , \\ C\left( {{t_p}} \right) = C\left( {{t_0}} \right) + \frac{{{h^\alpha }}}{{\Gamma \left( {\alpha + 1} \right)}}\sum\limits_{j = 0}^{p - 1} {{c_{j, p}}\left[ {\gamma S\left( {{t_j}} \right) - \left( {\omega + \mu } \right)C\left( {{t_j}} \right)} \right]} , \\ R\left( {{t_p}} \right) = R\left( {{t_0}} \right) + \frac{{{h^\alpha }}}{{\Gamma \left( {\alpha + 1} \right)}}\sum\limits_{j = 0}^{p - 1} {{c_{j, p}}\left[ {\eta I\left( {{t_j}} \right) + \varepsilon S\left( {{t_j}} \right) + \omega C\left( {{t_j}} \right) - \mu R\left( {{t_j}} \right)} \right]} , \\ M\left( {{t_p}} \right) = M\left( {{t_0}} \right) + \frac{{{h^\alpha }}}{{\Gamma \left( {\alpha + 1} \right)}}\sum\limits_{j = 0}^{p - 1} {{c_{j, p}}\left[ {{m_0} + \rho M\left( {{t_j}} \right) - \xi M\left( {{t_j}} \right)} \right]} . \\ \end{array}

    for p = 1, 2, \ldots, N where {c_{j, p}} = {\left({p - j} \right)^\alpha } - {\left({p - 1 - j} \right)^\alpha } . The fractional Euler method is employed for the subsequent numerical solutions of the fractional-order differential equations.

    In this section, we conduct numerical simulations to validate the findings presented in the preceding sections. The initial values are considered as I\left(0 \right) = 10500, {\text{ }}S\left(0 \right) = 1500, C\left(0 \right) = 750, {\text{ }}R\left(0 \right) = 750, and M\left(0 \right) = 1500 . Additionally, the parametric values are listed in Table 1. Obviously, the basic reproduction number is {{\text{R}}_0}{\text{ = 0}}{\text{.4}} < 1 . Figure 2 provides graphs of the evolution of (a) Ignorants, (b) Spreaders, (c) Calmness, (d) Removers, and (e) Media Reports in the rumor-free case for Eq (3.1), employing various values of the differentiation parameter \alpha = 0.6, 0.7, 0.8, 0.9, 1. With the decrease of the order \alpha , the convergence rate of the rumor propagation curve slows down and eventually tends to stabilize, which shows the memory effect. From the perspective of public opinion dissemination, it implies that such rumors tend to persist for an extended duration. Moreover, it demonstrates an anomalous propagation pattern at a lower order \alpha , characterized by an exponentially rapid growth in the initial phase as depicted in Figure 3(b). Meanwhile, the subsequent phase exhibits a deceleration. When the order is \alpha = 1 , the fractional-order differential equation turns into an integer-order one.

    Table 1.  Parameters under initial conditions 1.
    Parameter Value Parameter Value
    \Lambda 0.4 \omega 0.4
    \beta 0.1 \eta 0.2
    \delta 0.2 \varepsilon 0.5
    b 0.5 {m_0} 0.2
    k 0.5 \rho 0.2
    \xi 0.3 \mu 0.05
    \gamma 0.5

     | Show Table
    DownLoad: CSV
    Figure 2.  Dynamical behavior around the rumor-free equilibrium {E_0} .

    The initial values are also set to I\left(0 \right) = 10500, {\text{ }}S\left(0 \right) = 1500, {\text{ }}C\left(0 \right) = 750, {\text{ }}R\left(0 \right) = 750, and M\left(0 \right) = 1500 . Furthermore, the parametric values are provided in Table 2. The calculation result of the basic reproduction number is {{\text{R}}_0}{\text{ = 1}}{\text{.32}} > 1 . The results corresponding to the simulation of the rumors case are presented in Figure 3 using various values of the differentiation parameter \alpha = 0.6, 0.7, 0.8, 0.9, 1. The convergence rates of rumor propagation vary depending on the order value \alpha . The impact of the fractional order is observed through numerical simulation in Figure 3, from which it is demonstrated that reducing the value \alpha can enhance the stability range of model (3.1). The larger the order \alpha , the more rapidly the curve approaches stability in the rumor-permanence case. Therefore, it is imperative to consistently and promptly disseminate scientific knowledge while actively debunking rumors in order to effectively suppress rumor propagation.

    Table 2.  Parameters under initial conditions 2.
    Parameter Value Parameter Value
    \Lambda 1.2 \omega 0.4
    \beta 0.6 \eta 0.2
    \delta 0.5 \varepsilon 0.6
    b 1 {m_0} 0.6
    k 1 \rho 0.2
    \xi 0.3 \mu 0.3
    \gamma 0.5

     | Show Table
    DownLoad: CSV
    Figure 3.  Stability of the rumor-permanence equilibrium {E^*} .

    To illustrate the effectiveness of the media reports, for the basic reproduction number {R_0} > 1, {\text{ }}\alpha {\text{ = }}0.95 is established with the parametric values in Table 2. The media reports have an impact on Eq (3.1) governing fractional-order \alpha {\text{ = }}0.95 as illustrated in Figure 4. The peak of the Spreaders group decreases as the value of k increases, as shown in Figure 4(b). The statement suggests that there is a positive correlation between the level of media reports and the influence on rumor propagation. Therefore, employing appropriate media strategies can effectively suppress the rumor propagation.

    Figure 4.  Sample routes of I\left(t \right), S\left(t \right), C\left(t \right), R\left(t \right), and M\left(t \right) with different k for \alpha = 0.95 .

    To demonstrate the effectiveness of the fractional optimal control strategy, for the basic reproduction number {R_0} > 1, {\text{ }}\alpha {\text{ = }}0.95 is established with the parametric values in Table 2. Figure 5 shows the populations of I\left(t \right), S\left(t \right), C\left(t \right), R\left(t \right), and M\left(t \right) , with or without control mechanism. Compared with Eq (5.1), with or without control strategy, the populations of the Ignorants and the Spreaders are significantly influenced. This implies that implementing an effective control mechanism can significantly restrict the rumor diffusion.

    Figure 5.  Effect of different control strategies for \alpha = 0.95 .

    A Caputo sense-based ISCRM rumor propagation model incorporating media reports is proposed. The existence of positive solutions is rigorously established in this study, and the equilibrium solutions are determined for both the rumor-free equilibrium and rumor-permanence equilibrium, along with their local properties. The rumor-free equilibrium {E_0} is locally and globally asymptotically stable if {R_0} < 1 . The instability of the unique rumor-permanence equilibrium {E^*} exists when {R_0} > 1 , and for {l_1}{l_2} - {l_3} > 0, {\text{ }}{l_1}{l_2}{l_3} - l_1^2{l_3} - l_3^2 > 0 , it is locally asymptotically stable. The analytical research was validated by conducting numerical simulations, encompassing a range of fractional orders and initial compartment sizes. In the case of {R_0} < 1 , stability is rapidly achieved at lower fractional orders, whereas in the case of {R_0} > 1 , the attainment of stability is rapidly achieved at higher orders. The significance of fractional analysis has been underscored by the achievement of significantly improved and more realistic outcomes. Additionally, a comprehensive analysis of the impact of media reports and an optimal control strategy is presented. The effectiveness of rumor diffusion prevention can be significantly enhanced through the implementation of a control strategy and media reports, thereby mitigating the risk associated with rumor propagation. In future research, the investigation of time delay effects [46] and stochastic factors [47,48] could be expanded within the framework of complex networks in Caputo sense systems.

    The authors declare they have not used Artificial Intelligence (AI) tools in the creation of this article.

    The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.



    [1] Y. J. Chi, D. Wang, J. P. Wang, W. D. Yu, J. C. Yang, Long non-coding rna in the pathogenesis of cancers, Cells, 8 (2019), 1015. https://doi.org/10.3390/cells8091015 doi: 10.3390/cells8091015
    [2] S. Djebali, C. A. Davis, A. Merkel, A. Dobin, T. Lassmann, A. Mortazavi, et al., Landscape of transcription in human cells, Nature, 489 (2012), 101–108. https://doi.org/10.1038/nature11233 doi: 10.1038/nature11233
    [3] A. T. Willingham, A. P. Orth, S. Batalov, E. C. Peters, B. G. Wen, P. Aza-Blanc, et al., A strategy for probing the function of noncoding rnas finds a repressor of nfat, Science, 309 (2005), 1570–1573. https://doi.org/10.1126/science.1115901 doi: 10.1126/science.1115901
    [4] C. Xing, S. G. Sun, Z. Q. Yue, F. Bai, Role of lncrna lucat1 in cancer, Biomed. Pharmacother., 134 (2021), 111158. https://doi.org/10.1016/j.biopha.2020.111158 doi: 10.1016/j.biopha.2020.111158
    [5] L. Peng, M. Peng, B. Liao, G. H. Huang, W. B. Li, D. F. Xie, The advances and challenges of deep learning application in biological big data processing, Curr. Bioinf., 13 (2018), 352–359. https://doi.org/10.1163/9789004392533_041 doi: 10.1163/9789004392533_041
    [6] R. H. Wang, Y. Jiang, J. R. Jin, C. L. Yin, H. Q. Yu, F. S. Wang, et al., Deepbio: an automated and interpretable deep-learning platform for high-throughput biological sequence prediction, functional annotation and visualization analysis, Nucleic Acids Res., 51 (2023), 3017–3029. https://doi.org/10.1093/nar/gkad055 doi: 10.1093/nar/gkad055
    [7] L. H. Peng, J. W. Tan, W. Xiong, L. Zhang, Z. Wang, R. Y. Yuan, et al., Deciphering ligand-receptor-mediated intercellular communication based on ensemble deep learning and the joint scoring strategy from single-cell transcriptomic data, Comput. Biol. Med., 163 (2023), 107137. https://doi.org/10.1016/j.compbiomed.2023.107137 doi: 10.1016/j.compbiomed.2023.107137
    [8] W. Liu, Y. Yang, X. Lu, X. Z. Fu, R. Q. Sun, L. Yang, et al., Nsrgrn: a network structure refinement method for gene regulatory network inference, Briefings Bioinf., 24 (2023), bbad129. https://doi.org/10.1093/bib/bbad129 doi: 10.1093/bib/bbad129
    [9] J. C. Wang, Y. J. Chen, Q. Zou, Inferring gene regulatory network from single-cell transcriptomes with graph autoencoder model, PLos Genet., 19 (2023), e1010942. https://doi.org/10.1371/journal.pgen.1010942 doi: 10.1371/journal.pgen.1010942
    [10] L. Peng, C. Yang, L. Huang, X. Chen, X. Z. Fu, W. Liu, Rnmflp: predicting circrna-disease associations based on robust nonnegative matrix factorization and label propagation, Briefings Bioinf., 23 (2022), bbac155. https://doi.org/10.1093/bib/bbac155 doi: 10.1093/bib/bbac155
    [11] W. Liu, T. T. Tang, X. Lu, X. Z. Fu, Y. Yang, L. Peng, Mpclcda: predicting circrna-disease associations by using automatically selected meta-path and contrastive learning, Briefings Bioinf., 24 (2023), bbad227. https://doi.org/10.1093/bib/bbad227 doi: 10.1093/bib/bbad227
    [12] L. L. Zhuo, S. Y. Pan, J. Li, X. Z. Fu, Predicting mirna-lncrna interactions on plant datasets based on bipartite network embedding method, Methods, 207 (2022), 97–102. https://doi.org/10.1016/j.ymeth.2022.09.002 doi: 10.1016/j.ymeth.2022.09.002
    [13] Z. C. Zhou, Z. Y. Du, J. H. Wei, L. L. Zhuo, S. Y. Pan, X. Z. Fu, et al., Mham-npi: Predicting ncrna-protein interactions based on multi-head attention mechanism, Comput. Biol. Med., 163 (2023), 107143. https://doi.org/10.1016/j.compbiomed.2023.107143 doi: 10.1016/j.compbiomed.2023.107143
    [14] X. Chen, L. Wang, J. Qu, N. N. Guan, J. Q. Li, Predicting mirna-disease association based on inductive matrix completion, Bioinformatics, 34 (2018), 4256–4265. https://doi.org/10.1093/bioinformatics/bty503 doi: 10.1093/bioinformatics/bty503
    [15] X. Chen, D. Xie, L. Wang, Q. Zhao, Z. H. You, H. Liu, Bnpmda: Bipartite network projection for mirna-disease association prediction, Bioinformatics, 34 (2018), 3178–3186. https://doi.org/10.1093/bioinformatics/bty333 doi: 10.1093/bioinformatics/bty333
    [16] L. Huang, L. Zhang, X. Chen, Updated review of advances in micrornas and complex diseases: towards systematic evaluation of computational models, Briefings Bioinf., 23 (2022), bbac407. https://doi.org/10.1093/bib/bbac407 doi: 10.1093/bib/bbac407
    [17] C. C. Wang, C. C. Zhu, X. Chen, Ensemble of kernel ridge regression-based small molecule-mirna association prediction in human disease, Briefings Bioinf., 23 (2022), bbab431. https://doi.org/10.1093/bib/bbab431 doi: 10.1093/bib/bbab431
    [18] Z. J. Li, Y. X. Zhang, Y. Bai, X. H. Xie, L. J. Zeng, Imc-mda: Prediction of mirna-disease association based on induction matrix completion, Math. Biosci. Eng., 20 (2023), 10659–10674. https://doi.org/10.3934/mbe.2023471 doi: 10.3934/mbe.2023471
    [19] Q. Qu, X. Chen, B. Ning, X. Zhang, H. Nie, L. Zeng, et al., Prediction of mirna-disease associations by neural network-based deep matrix factorization, Methods, 212 (2023), 1–9. https://doi.org/10.1016/j.ymeth.2023.02.003 doi: 10.1016/j.ymeth.2023.02.003
    [20] L. Zhang, C. C. Wang, X. Chen, Predicting drug-target binding affinity through molecule representation block based on multi-head attention and skip connection, Briefings Bioinf., 23 (2022), bbac468. https://doi.org/10.1093/bib/bbac468 doi: 10.1093/bib/bbac468
    [21] L. Katusiime, Covid-19 and the effect of central bank intervention on exchange rate volatility in developing countries: The case of uganda, National Accounting Rev., 5 (2023), 23–37. https://doi.org/10.3934/NAR.2023002 doi: 10.3934/NAR.2023002
    [22] L. Grassini, Statistical features and economic impact of Covid-19, National Accounting Rev., 5 (2023), 38–40. https://doi.org/10.3934/NAR.2023003 doi: 10.3934/NAR.2023003
    [23] Z. Y. Bao, Z. Yang, Z. Huang, Y. R. Zhou, Q. H. Cui, D. Dong, Lncrnadisease 2.0: an updated database of long non-coding rna-associated diseases, Nucleic Acids Res., 47 (2019), D1034–D1037. https://doi.org/10.1093/nar/gky905 doi: 10.1093/nar/gky905
    [24] S. W. Ning, J. Z. Zhang, P. Wang, H. Zhi, J. J. Wang, Y. Liu, et al., Lnc2cancer: a manually curated database of experimentally supported lncrnas associated with various human cancers, Nucleic Acids Res., 44 (2016), D980–D985. https://doi.org/10.1093/nar/gkv1094 doi: 10.1093/nar/gkv1094
    [25] X. Chen, L. Huang, Computational model for disease research, Briefings Bioinf., 24 (2023), bbac615. https://doi.org/10.1093/bib/bbac615 doi: 10.1093/bib/bbac615
    [26] K. Albitar, K. Hussainey, Sustainability, environmental responsibility and innovation, Green Finance, 5 (2023), 85–88. https://doi.org/10.3934/GF.2023004 doi: 10.3934/GF.2023004
    [27] G. Desalegn, Insuring a greener future: How green insurance drives investment in sustainable projects in developing countries, Green Finance, 5 (2023), 195–210. https://doi.org/10.3934/GF.2023008 doi: 10.3934/GF.2023008
    [28] Y. Liang, Z. H. Zhang, N. N. Liu, Y. N. Wu, C. L. Gu, Y. L. Wang, Magcnse: predicting lncrna-disease associations using multi-view attention graph convolutional network and stacking ensemble model, BMC Bioinf., 23 (2022). https://doi.org/10.1186/s12859-022-04715-w doi: 10.1186/s12859-022-04715-w
    [29] Y. Kim, M. Lee, Deep learning approaches for lncrna-mediated mechanisms: A comprehensive review of recent developments, Int. J. Mol. Sci., 24 (2023), 10299. https://doi.org/10.3390/ijms241210299 doi: 10.3390/ijms241210299
    [30] Z. Q. Zhang, J. L. Xu, Y. N. Wu, N. N. Liu, Y. L. Wang, Y. Liang, Capsnet-lda: predicting lncrna-disease associations using attention mechanism and capsule network based on multi-view data, Briefings Bioinf., 24 (2022), bbac531. https://doi.org/10.1093/bib/bbac531 doi: 10.1093/bib/bbac531
    [31] N. Dwarika, The risk-return relationship and volatility feedback in south africa: a comparative analysis of the parametric and nonparametric bayesian approach, Quant. Finance Econ., 7 (2023), 119–146. https://doi.org/10.3934/QFE.2023007 doi: 10.3934/QFE.2023007
    [32] N. Dwarika, Asset pricing models in south africa: A comparative of regression analysis and the bayesian approach, Data Sci. Finance Econ., 3 (2023), 55–75. https://doi.org/10.3934/DSFE.2023004 doi: 10.3934/DSFE.2023004
    [33] Y. Q. Lin, X. J. Chen, H. Y. Lan, Analysis and prediction of american economy under different government policy based on stepwise regression and support vector machine modelling, Data Sci. Finance Econ., 3 (2023), 1–13. https://doi.org/10.3934/DSFE.2023001 doi: 10.3934/DSFE.2023001
    [34] N. Sheng, L. Huang, Y. T. Lu, H. Wang, L. L. Yang, L. Gao, et al., Data resources and computational methods for lncrna-disease association prediction, Comput. Biol. Med., 153 (2023), 106527. https://doi.org/10.1016/j.compbiomed.2022.106527 doi: 10.1016/j.compbiomed.2022.106527
    [35] J. H. Wei, L. L. Zhuo, S. Y. Pan, X. Z. Lian, X. J. Yao, X. Z. Fu, Headtailtransfer: An efficient sampling method to improve the performance of graph neural network method in predicting sparse ncrna-protein interactions, Comput. Biol. Med., 157 (2023), 106783. https://doi.org/10.1016/j.compbiomed.2023.106783 doi: 10.1016/j.compbiomed.2023.106783
    [36] P. Xuan, S. X. Pan, T. G. Zhang, Y. Liu, H. Sun, Graph convolutional network and convolutional neural network based method for predicting lncrna-disease associations, Cells, 8 (2019), 1012. https://doi.org/10.3390/cells8091012 doi: 10.3390/cells8091012
    [37] M. F. Leung, A. Jawaid, S. W. Ip, C. H. Kwok, S. Yan, A portfolio recommendation system based on machine learning and big data analytics, Data Sci. Finance Econ., 3 (2023), 152–165. https://doi.org/10.3934/DSFE.2023009 doi: 10.3934/DSFE.2023009
    [38] Q. W. Wu, J. F. Xia, J. C. Ni, C. H. Zheng, Gaerf: predicting lncrna-disease associations by graph auto-encoder and random forest, Briefings Bioinf., 22 (2021), bbaa391. https://doi.org/10.1093/bib/bbaa391 doi: 10.1093/bib/bbaa391
    [39] N. Sheng, L. Huang, Y. Wang, J. Zhao, P. Xuan, L. Gao, et al., Multi-channel graph attention autoencoders for disease-related lncrnas prediction, Briefings Bioinf., 23 (2022), bbab604. https://doi.org/10.1093/bib/bbab604 doi: 10.1093/bib/bbab604
    [40] L. Peng, C. Yang, Y. F. Chen, W. Liu, Predicting circrna-disease associations via feature convolution learning with heterogeneous graph attention network, IEEE J. Biomed. Health. Inf., 27 (2023), 3072–3082. https://doi.org/10.1109/JBHI.2023.3260863. doi: 10.1109/JBHI.2023.3260863
    [41] X. Liu, C. Z. Song, F. Huang, H. T. Fu, W. J. Xiao, W. Zhang, Graphcdr: a graph neural network method with contrastive learning for cancer drug response prediction, Briefings Bioinf., 23 (2022), bbab457. https://doi.org/10.1093/bib/bbab457 doi: 10.1093/bib/bbab457
    [42] G. Y. Fu, J. Wang, C. Domeniconi, G. X. Yu, Matrix factorization-based data fusion for the prediction of lncrna–disease associations, Bioinformatics, 34 (2018), 1529–1537. https://doi.org/10.1093/bioinformatics/btx794 doi: 10.1093/bioinformatics/btx794
    [43] Z. Y. Lu, K. B. Cohen, L. Hunter, Generif quality assurance as summary revision, in Biocomputing 2007, World Scientific, (2007), 269–280. https://doi.org/10.1142/9789812772435_0026
    [44] J. H. Li, S. Liu, H. Zhou, L. H. Qu, J. H. Yang, starBase v2. 0: decoding miRNA-ceRNA, miRNA-ncRNA and protein–RNA interaction networks from large-scale CLIP-Seq data, Nucleic Acids Res., 42 (2014), D92–D97. https://doi.org/10.1093/nar/gkt1248 doi: 10.1093/nar/gkt1248
    [45] W. Lan, Y. Dong, Q. F. Chen, R. Q. Zheng, J. Liu, Y. Pan, et al., Kgancda: predicting circrna-disease associations based on knowledge graph attention network, Briefings Bioinf., 23 (2022), bbab494. https://doi.org/10.1093/bib/bbab494 doi: 10.1093/bib/bbab494
    [46] Z. H. Guo, Z. H. You, D. S. Huang, H. C. Yi, Z. H. Chen, Y. B. Wang, A learning based framework for diverse biomolecule relationship prediction in molecular association network, Commun. Biol., 3 (2020). https://doi.org/10.1038/s42003-020-0858-8 doi: 10.1038/s42003-020-0858-8
    [47] D. Wang, J. Wang, M. Lu, F. Song, Q. H. Cui, Inferring the human microRNA functional similarity and functional network based on microRNA-associated diseases, Bioinformatics, 26 (2010), 1644–1650. https://doi.org/10.1093/bioinformatics/btq241 doi: 10.1093/bioinformatics/btq241
    [48] X. Chen, Predicting lncRNA-disease associations and constructing lncRNA functional similarity network based on the information of miRNA, Sci. Rep., 5 (2015), 13186. https://doi.org/10.1038/srep13186 doi: 10.1038/srep13186
    [49] X. Chen, G. Y. Yan, Novel human lncrna-disease association inference based on lncrna expression profiles, Bioinformatics, 29 (2013), 2617–2624. https://doi.org/10.1093/bioinformatics/btt426 doi: 10.1093/bioinformatics/btt426
    [50] D. Anderson, U. Ulrych, Accelerated american option pricing with deep neural networks, Quant. Finance Econ., 7 (2023), 207–228. https://doi.org/10.3934/QFE.2023011 doi: 10.3934/QFE.2023011
    [51] T. N. Kipf, M. Welling, Semi-supervised classification with graph convolutional networks, preprint, arXiv: 1609.02907. https://doi.org/10.48550/arXiv.1609.02907
    [52] L. Peng, Y. Tu, L. Huang, Y. Li, X. Z. Fu, X. Chen, Daestb: inferring associations of small molecule–mirna via a scalable tree boosting model based on deep autoencoder, Briefings Bioinf., 23 (2022), bbac478. https://doi.org/10.1093/bib/bbac478 doi: 10.1093/bib/bbac478
    [53] Z. Y. Chu, S. C. Liu, W. Zhang, Hierarchical graph representation learning for the prediction of drug-target binding affinity, Inf. Sci., 613 (2022), 507–523. https://doi.org/10.1016/j.ins.2022.09.043 doi: 10.1016/j.ins.2022.09.043
    [54] M. Chen, Z. W. Wei, Z. F. Huang, B. L. Ding, Y. L. Li, Simple and deep graph convolutional networks, in Proceedings of the 37th International Conference on Machine Learning, PMLR, (2020), 1725–1735.
    [55] X. Chen, Katzlda: Katz measure for the lncrna-disease association prediction, Sci. Rep., 5 (2015), 16840. https://doi.org/10.1038/srep16840 doi: 10.1038/srep16840
    [56] C. Q. Lu, M. Y. Yang, F. Luo, F. X. Wu, M. Li, Y. Pan, et al., Prediction of lncrna–disease associations based on inductive matrix completion, Bioinformatics, 34 (2018), 3357–3364. https://doi.org/10.1093/bioinformatics/bty327 doi: 10.1093/bioinformatics/bty327
    [57] X. M. Wu, W. Lan, Q. F. Chen, Y. Dong, J. Liu, W. Peng, Inferring LncRNA-disease associations based on graph autoencoder matrix completion, Comput. Biol. Chem., 87 (2020), 107282. https://doi.org/10.1016/j.compbiolchem.2020.107282 doi: 10.1016/j.compbiolchem.2020.107282
    [58] M. Zeng, C. Q. Lu, Z. H. Fei, F. X. Wu, Y. H. Li, J. X. Wang, et al., Dmflda: a deep learning framework for predicting lncrna–disease associations, IEEE/ACM Trans. Comput. Biol. Bioinf., 18 (2020), 2353–2363. https://doi.org/10.1109/TCBB.2020.2983958. doi: 10.1109/TCBB.2020.2983958
    [59] R. Zhu, Y. Wang, J. X. Liu, L. Y. Dai, Ipcarf: improving lncrna-disease association prediction using incremental principal component analysis feature selection and a random forest classifier, BMC Bioinf., 22 (2021). https://doi.org/10.1186/s12859-021-04104-9 doi: 10.1186/s12859-021-04104-9
    [60] Y. S. Sun, Z. Zhao, Z. N. Yang, F. Xu, H. J. Lu, Z. Y. Zhu, et al., Risk factors and preventions of breast cancer, Int. J. Biol. Sci., 13 (2017), 1387–1397. https://doi.org/10.7150/ijbs.21635 doi: 10.7150/ijbs.21635
    [61] H. Jin, W. Du, W. T. Huang, J. J. Yan, Q. Tang, Y. B. Chen, et al., lncRNA and breast cancer: Progress from identifying mechanisms to challenges and opportunities of clinical treatment, Mol. Ther.–Nucleic Acids, 25 (2021), 613–637. https://doi.org/10.1016/j.omtn.2021.08.005 doi: 10.1016/j.omtn.2021.08.005
    [62] J. J. Xu, M. S. Hu, Y. Gao, Y. S. Wang, X. N. Yuan, Y. Yang, et al., Lncrna mir17hg suppresses breast cancer proliferation and migration as cerna to target fam135a by sponging mir-454-3p, Mol. Biotechnol., 65 (2023), 2071–2085. https://doi.org/10.1007/s12033-023-00706-1 doi: 10.1007/s12033-023-00706-1
    [63] K. X. Lou, Z. H. Li, P. Wang, Z. Liu, Y. Chen, X. L. Wang, et al., Long non-coding rna bancr indicates poor prognosis for breast cancer and promotes cell proliferation and invasion, Eur. Rev. Med. Pharmacol. Sci., 22 (2018), 1358–1365.
    [64] F. Bray, J. Ferlay, I. Soerjomataram, R. L. Siegel, L. A. Torre, A. Jemal, Global cancer statistics 2018: Globocan estimates of incidence and mortality worldwide for 36 cancers in 185 countries, CA: Cancer J. Clinicians, 68 (2018), 394–424. https://doi.org/10.3322/caac.21492 doi: 10.3322/caac.21492
    [65] Z. W. Wang, Y. Y. Jin, H. T. Ren, X. L. Ma, B. F. Wang, Y. L. Wang, Downregulation of the long non-coding RNA TUSC7 promotes NSCLC cell proliferation and correlates with poor prognosis, Am. J. Transl. Res., 8 (2016), 680–687.
    [66] H. P. Deng, L. Chen, T. Fan, B. Zhang, Y. Xu, Q. Geng, Long non-coding rna hottip promotes tumor growth and inhibits cell apoptosis in lung cancer, Cell. Mol. Biol., 61 (2015), 34–40.
    [67] H. Sung, J. Ferlay, R. L. Siegel, M. Laversanne, I. Soerjomataram, A. Jemal, et al., Global cancer statistics 2020: Globocan estimates of incidence and mortality worldwide for 36 cancers in 185 countries. CA: Cancer J. Clinicians, 71 (2021), 209–249. https://doi.org/10.3322/caac.21660 doi: 10.3322/caac.21660
    [68] J. Q. Wang, L. P. Su, X. H. Chen, P. Li, Q. Cai, B. Q. Yu, et al., MALAT1 promotes cell proliferation in gastric cancer by recruiting SF2/ASF, Biomed. Pharmacother., 68 (2014), 557–564. https://doi.org/10.1016/j.biopha.2014.04.007 doi: 10.1016/j.biopha.2014.04.007
    [69] L. Ma, Y. J. Zhou, X. J. Luo, H. Gao, X. B. Deng, Y. J. Jiang, Long non-coding RNA XIST promotes cell growth and invasion through regulating miR-497/MACC1 axis in gastric cancer, Oncotarget, 8 (2017), 4125–4135. https://doi.org/10.18632/oncotarget.13670 doi: 10.18632/oncotarget.13670
    [70] H. T. Fu, F. Huang, X. Liu, Y. Qiu, W. Zhang, Mvgcn: data integration through multi-view graph convolutional network for predicting links in biomedical bipartite networks, Bioinformatics, 38 (2022), 426–434. https://doi.org/10.1093/bioinformatics/btab651 doi: 10.1093/bioinformatics/btab651
  • This article has been cited by:

    1. Hong Li, Zhantao Li, Chaohong Pan, Jie Song, Mingji Zhang, Cubic-like Features of I–V Relations via Classical Poisson–Nernst–Planck Systems Under Relaxed Electroneutrality Boundary Conditions, 2024, 13, 2075-1680, 790, 10.3390/axioms13110790
    2. Xiangshuo Liu, Jie Song, Lijun Zhang, Mingji Zhang, Roles Played by Critical Potentials in the Study of Poisson–Nernst–Planck Models with Steric Effects Under Relaxed Neutral Boundary Conditions, 2025, 14, 2075-1680, 69, 10.3390/axioms14010069
  • Reader Comments
  • © 2024 the Author(s), licensee AIMS Press. This is an open access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0)
通讯作者: 陈斌, bchen63@163.com
  • 1. 

    沈阳化工大学材料科学与工程学院 沈阳 110142

  1. 本站搜索
  2. 百度学术搜索
  3. 万方数据库搜索
  4. CNKI搜索

Metrics

Article views(1388) PDF downloads(166) Cited by(0)

Figures and Tables

Figures(10)  /  Tables(6)

Other Articles By Authors

/

DownLoad:  Full-Size Img  PowerPoint
Return
Return

Catalog