
In recent years, the number of fintech startups has increased substantially, especially those that offer financial services in areas considered as traditional advantage of commercial banks such as payment, deposit or lending services. This paper examines the cooperation in the provision of financial services between banks and Fintech from consumers' perspective, which is the customers' adoption of financial services offered by the partnership between banks and Fintechs. The study employs SEM utilizing data of 234 customers using financial services in Vietnam to identify the factors influence their behaviors and intention to use financial services offered by banks and fintechs partnerships. The findings indicate that perceived usefulness, brand image, user innovativeness and trust have positive effects on customers' attitude and consequently on their intention to use. However, there is no evidence regarding the impact of perceived ease of use and perceived risks on behavior and intention to use of customers. The findings also provide more evidence on the factors that influence the intention of customers' technology acceptance in the traditional Technology Acceptance Model (TAM).
Citation: Yen H. Hoang, Duong T.T. Nguyen, Linh H.T. Tran, Nhung T.H. Nguyen, Ngoc B. Vu. Customers' adoption of financial services offered by banks and fintechs partnerships: evidence of a transitional economy[J]. Data Science in Finance and Economics, 2021, 1(1): 77-95. doi: 10.3934/DSFE.2021005
[1] | Ting Guo, Zhipeng Qiu . The effects of CTL immune response on HIV infection model with potent therapy, latently infected cells and cell-to-cell viral transmission. Mathematical Biosciences and Engineering, 2019, 16(6): 6822-6841. doi: 10.3934/mbe.2019341 |
[2] | A. M. Elaiw, A. S. Shflot, A. D. Hobiny . Stability analysis of general delayed HTLV-I dynamics model with mitosis and CTL immunity. Mathematical Biosciences and Engineering, 2022, 19(12): 12693-12729. doi: 10.3934/mbe.2022593 |
[3] | Xuejuan Lu, Lulu Hui, Shengqiang Liu, Jia Li . A mathematical model of HTLV-I infection with two time delays. Mathematical Biosciences and Engineering, 2015, 12(3): 431-449. doi: 10.3934/mbe.2015.12.431 |
[4] | Yan Wang, Minmin Lu, Daqing Jiang . Viral dynamics of a latent HIV infection model with Beddington-DeAngelis incidence function, B-cell immune response and multiple delays. Mathematical Biosciences and Engineering, 2021, 18(1): 274-299. doi: 10.3934/mbe.2021014 |
[5] | Cameron Browne . Immune response in virus model structured by cell infection-age. Mathematical Biosciences and Engineering, 2016, 13(5): 887-909. doi: 10.3934/mbe.2016022 |
[6] | Jiawei Deng, Ping Jiang, Hongying Shu . Viral infection dynamics with mitosis, intracellular delays and immune response. Mathematical Biosciences and Engineering, 2023, 20(2): 2937-2963. doi: 10.3934/mbe.2023139 |
[7] | Gesham Magombedze, Winston Garira, Eddie Mwenje . Modelling the human immune response mechanisms to mycobacterium tuberculosis infection in the lungs. Mathematical Biosciences and Engineering, 2006, 3(4): 661-682. doi: 10.3934/mbe.2006.3.661 |
[8] | Yu Yang, Gang Huang, Yueping Dong . Stability and Hopf bifurcation of an HIV infection model with two time delays. Mathematical Biosciences and Engineering, 2023, 20(2): 1938-1959. doi: 10.3934/mbe.2023089 |
[9] | Yan Wang, Tingting Zhao, Jun Liu . Viral dynamics of an HIV stochastic model with cell-to-cell infection, CTL immune response and distributed delays. Mathematical Biosciences and Engineering, 2019, 16(6): 7126-7154. doi: 10.3934/mbe.2019358 |
[10] | Xiaohong Tian, Rui Xu, Jiazhe Lin . Mathematical analysis of an age-structured HIV-1 infection model with CTL immune response. Mathematical Biosciences and Engineering, 2019, 16(6): 7850-7882. doi: 10.3934/mbe.2019395 |
In recent years, the number of fintech startups has increased substantially, especially those that offer financial services in areas considered as traditional advantage of commercial banks such as payment, deposit or lending services. This paper examines the cooperation in the provision of financial services between banks and Fintech from consumers' perspective, which is the customers' adoption of financial services offered by the partnership between banks and Fintechs. The study employs SEM utilizing data of 234 customers using financial services in Vietnam to identify the factors influence their behaviors and intention to use financial services offered by banks and fintechs partnerships. The findings indicate that perceived usefulness, brand image, user innovativeness and trust have positive effects on customers' attitude and consequently on their intention to use. However, there is no evidence regarding the impact of perceived ease of use and perceived risks on behavior and intention to use of customers. The findings also provide more evidence on the factors that influence the intention of customers' technology acceptance in the traditional Technology Acceptance Model (TAM).
Microorganisms continuously have to contend with environmental changes in nature, and so they have evolved to accordingly adapt their physiology to cope with this unsteadiness. This is done by reorganizing the gene expression machinery, which is accomplished by dynamically allocating resources to different cellular functions. Microorganisms like bacteria exhibit great genetic variability, which is the main driver of natural selection, a phenomenon that depends on the continuous mutations in living populations. For this reason, they have achieved highly optimized resource management strategies. Such is the case of E. Coli: For specific environmental conditions, they seek to maximize their growth rate, which is a naturally evolved characteristic of this bacterium, and well known in the scientific community [1]. In this regard, optimality of bacterial organisms has been a central hypothesis in several research allocation studies [2].
Nevertheless, most of previous works in the literature consider the resource allocation problem in steady-state, without taking into account the changing conditions of natural environments [3]. This motivated a series of works with a dynamical-oriented approach to the problem [4,5], that aimed to shed light on how bacteria tune their allocation mechanisms when facing changes on the nutrient concentration of the medium. Using Optimal Control theory, they investigate how their internal pathways can be dynamically readjusted so as to maximize their growth rate, obtaining different strategies that can be related to well known natural regulation mechanisms (such as the ppGpp-mediated sensing of the pool of aminoacids). The approach is based on a widely used modeling technique in systems biology: The so-called coarse-grained self-replicator models, used in bacterial growth representations for their simplicity and their capacity to reproduce observed experimental behaviors [6]. Although this kind of single-cell models are somewhat limited when predicting complex phenomena, they can accurately account for bacterial culture growth laws under the right assumptions (e.g., homogeneity of the culture) [7].
From an industry point of view, a natural question triggered by these studies is: How can we divert the natural allocation strategies of bacteria to improve current biotechnological production schemes? Such is the case of the artificial synthesis of metabolites or proteins of interest. The synthesis of such compounds is highly relevant for its wide range of applications: Production of antitumor agents, insulin, antibiotics, immunosuppressive agents and insecticides, among others [8,9]. Motivated by the increasing understanding of the biosynthetic properties of certain microorganisms, research on this area can potentially lead to more efficient and sustainable production schemes. This is the matter addressed in recent work using a strain of Escherichia coli that includes an artificially engineered heterologous pathway for the production of a certain metabolite of interest [10,11,12]. In this approach, a control loop is developed through a bacterial growth switch that allows to externally modify the natural resource allocation decision [13]. The mechanism is implemented by re-engineering the transcriptional control of the expression of RNA polymerase, a key component of the gene expression machinery. This way, it is possible to optimize the productivity of the bioprocess by channeling resources into this new heterologous pathway.
At the same time, the synthesis of these metabolites draws resources from the native pathways used for producing biomass in bacteria, thus leading to an inherent compromise between these two objectives. One possible approach to this trade-off is to model it through different cost functions, thus obtaining multi-objective optimization problems. This is the case of [14], where the authors aim to maximize the production of a metabolite while minimizing the genetic burden caused by pathway expression. In contrast to this method, the work of [10] models the main trade-offs behind the process through a single decision parameter, which considerably reduces the complexity of the optimization problem.
In a similar vein, we address a classic production scheme: The Continuous Stirred-Tank Reactor (CSTR). While resource allocation in bacteria has been vastly studied in constant environments (e.g., under the assumption that there is always enough substrate in the medium), how this goes in continuous bioreactors is not trivial, since a feedback occurs from the physiology of the cell to the environmental conditions given by the interaction bacteria-medium [15]. Examples of self-replicator models in continuous bioreactors can be found in some recent works: In [16], authors use a coarse-grained self-replicator kinetic model of Saccharomyces cerevisiae's in a continuous bioreactor to investigate the trade-off between respiratory and fermentative metabolism, showing that optimal strategies are ‘pure’ metabolic strategies (e.g., either respiration of fermentation, but not respiro-fermentation). Likewise, it is rather classical in the continuous bioreactor scheme to maximize a certain performance measure (i.e., biomass production) in terms of the operational parameters related to the setup, such as dilution rate and/or concentration of the substrate inflow [17,18,19,20,21]. However, incorporating the aforementioned external control—that can disrupt the natural allocation process of the whole culture—provides an extra degree of freedom, which can, in turn, contribute to further improve the classical production scheme.
Based on the presented works, we show novel results addressing the problem of bacterial resource allocation in the CSTR framework. Our approach is based on a coarse-grained self-replicator dynamical model that accounts for the microbial culture growth inside the continuous bioreactor, and incorporates the external allocation control previously described. The novelty of the approach lies in the combination of the resource allocation control scheme, and the capacity to regulate the bacterial growth rate through the dilution rate of the continuous bioreactor. Further on, we study its asymptotic behavior using dynamical systems theory, and we provide conditions for the persistence of the bacterial population. The analysis is carried out by studying the local and global behavior of its limiting system, and relating its convergence to the original model through the theory of asymptotically autonomous systems [22]. Then, we pose the problem of maximizing the synthesis of the metabolite of interest during a fixed interval of time in terms of the resource allocation decision. The latter is expressed as an OCP (Optimal Control Problem) which is then explored using PMP (Pontryagin's Maximum Principle) [23]. We analyze the solution of the problem and propose a sub-optimal control strategy given by a constant allocation decision, which eventually takes the system to the optimal steady-state production regime. On this basis, we study and compare the two most significant steady-state production objectives in CSTRs: Biomass production and metabolite production. For this last purpose, and in addition to the allocation parameter, we control the constant volumetric flow of the bioreactor (or dilution rate), and we analyze the results through a numerical approach. The resulting two-dimensional optimization problem is defined in terms of Michaelis-Menten kinetics with the parameter values of [4], and taking into account the constraints for the existence of the equilibrium of interest.
In this section, we define the coarse-grained self-replicator model including the continuous bioreactor scheme and the allocation parameter. We consider a growing bacterial population in a CSTR bioreactor of constant volume Vext [L]. The self-replicating system that models the culture is composed of the metabolic machinery M (transporters, enzymes...) and the gene expression machinery R (RNA polymerase, ribosomes...), both responsible of the cell growth. The validity of this single-cell model as a representation of a growing bacterial culture depends on a number of simplifying assumptions, one of them being that individual cells share the same macromolecular composition. For the production of the metabolite of interest X, we consider the artificially engineered pathway introduced in [10] (Figure 1).
The model represents three chemical macroreactions,
SVM⟶P |
PVR⟶αR+(1−α)M |
PVX⟶X |
The first reaction is catalyzed by M and describes the transformation of external substrate S into precursor metabolites P at a rate VM. The second one represents the conversion of precursors into macromolecules R and M and is catalyzed by R, at a rate VR. Finally, the third reaction describes the transformation of precursors P into product X at a rate VX, and catalyzed by M. The natural resource allocation parameter is modeled through the dimensionless parameter α∈[0,1], that represents the proportion of precursors allocated to the gene expression machinery R, while 1−α indicates that of the metabolic machinery M. The model describes all mass quantities S, P, M, R and X in grams, and the rates in grams per hour. A constant volumetric flow rate F [L h−1] produces an inflow of fresh medium rich in substrate, and an outflow of bacterial culture and metabolites [24]. As already stated, we include in our scheme the growth switch described in [13] to externally affect the allocation decision α by varying the inducer concentration in the medium. The mechanism is modeled as
u(t)=I(t)α(t),u∈[0,1] | (2.1) |
where I is the external signal, which acts independently from the allocation parameter α. The agammaegated form of Eq (2.1) accounts for the fact that the synthetic switch is capable of adjusting bacterial growth between zero (by setting I=0, which yields u=0) and the maximal growth rate supported by the medium (given by I=1, which yields u=α). In this work, we limit the analysis to calculating the control input u, without decoupling both individual controls I and α (refer to Section 5 in [10] for more details on the implementation of this external controller). Then, we write the system of differential equations describing the evolution of the mass of each component,
{˙S=VSin−VM−VSout,˙P=VM−VR−VX−VPout,˙M=(1−u)VR−VMout,˙R=uVR−VRout,˙X=VX−VXout. |
The inflow/outflow rates are defined as
VSout=DS,VPout=DP,VMout=DM,VRout=DR,VXout=DX,VSin=Fsin, |
where sin [g L−1] is the nutrient concentration of the inflow of fresh medium, and D [h−1] the dilution rate defined as
D≐FVext. |
We define the volume of the cell population V [L] as
V≐β(M+R), | (2.2) |
where β [L g−1] is the inverse of the cytoplasmic density. The above definition is based on the assumption that the cytoplasmic density of the cells is constant for the whole culture, and it also takes into account the experimental results showing that macromolecules are responsible for most of the biomass in microbial cells [25]. Thus, the mass of precursors P is excluded from the volume V as a simplifying assumption. Then, we express the quantities of the system as concentrations with respect to the volumes,
p≐PV,r≐RV,m≐MV,s≐SVext,x≐XVext, | (2.3) |
where p, r and m [g L−1] are intracellular concentrations of precursor metabolites, components of the gene expression machinery and metabolic enzymes respectively; and s and x [g L−1] the extracellular concentrations of substrate and metabolite. It is worth stressing that intracellular concentrations are defined with respect to the bacterial volume V, while extracellular concentrations with respect to the volume of the bioreactor Vext. Then, using definition (2.2), we obtain that m+r=1/β. We define the growth rate of the bacterial population μ [h−1] as the relative variation of cell volume ˙V/V without considering the effect of the outflowing biomass. Replacing with concentrations leads to the system
S:{˙s=D(sin−s)−vM(s,m)VVext,˙p=vM(s,m)−vR(p,r)−vX(p,m)−μ(p,r)p,˙r=uvR(p,r)−μ(p,r)r,˙m=(1−u)vR(p,r)−μ(p,r)m,˙x=vX(p,m)VVext−Dx,˙V=(μ(p,r)−D)V, | (2.4) |
where vM(s,m), vR(p,r) and vX(p,m) [g L−1 h−1] are the mass fluxes per unit volume obtained from dividing the rates VM, VR and VX by V, function of the concentrations of system S. In this new system, the growth rate becomes
μ(p,r)≐˙VV|F=0=˙M+˙RM+R|F=0=βvR(p,r) |
showing that the bacterial growth rate is proportional to the macromolecule synthesis rate. We propose a change of variables that simplifies the expressions of the system
ˆs=βs,ˆp=βp,ˆr=βr,ˆm=βm,ˆx=βx,ˆV=VVext, |
which yields the relation
ˆr+ˆm=1, | (2.5) |
and we define the non-dimensional synthesis rates
ˆvM(ˆs,ˆm)=βvM(s,m),ˆvR(ˆp,ˆr)=βvR(p,r),ˆvX(ˆp,ˆm)=βvX(p,m), | (2.6) |
and the non-dimensional substrate concentration ˆsin=βsin. Then, dropping all hats yields the following system
S1:{˙s=D(sin−s)−vM(s,1−r)V,˙p=vM(s,1−r)−vX(p,1−r)−μ(p,r)(p+1),˙r=(u−r)μ(p,r),˙x=vX(p,1−r)V−Dx,˙V=(μ(p,r)−D)V, | (2.7) |
where the dynamical expression of m has been omitted since it can be computed from r, as shown in Eq (2.5). It can also be seen that both concentrations m and r are limited to the interval [0,1] due to physical constraints from the relation in Eq (2.5). For this latter, and due to the nature of the reactions involved in the studied problem, we will make some assumptions on the synthesis rates vM(s,m), vR(p,r) and vX(p,m):
Assumption 1. Functions vM(s,m), vR(p,r) and vX(p,m) meet
● vi(y,z):R+×[0,1]→R+
● vi(y,z) continuously differentiable w.r.t. both variables
● vi(0,z)=vi(y,0)=0
● vi(⋅) strictly monotonically increasing:
∂vi∂y(y,z)>0,∀(y,z)∈R>0×(0,1],∂vi∂z(y,z)>0,∀(y,z)∈R>0×(0,1] |
● vi(⋅) bounded w.r.t. y: limy→∞vi(y,z)=vi,max(z).
Assumption 1 encompasses all general monotone increasing kinetics models used in biological models, such as Michaelis-Menten or Hill equation-based kinetics [26]. In this first work, we will focus on a particular kind of systems where the synthesis rate related to the metabolite production depends on the growth rate:
Assumption 2. For r∈(0,1), the metabolite synthesis rate vX(p,1−r) can be expressed in terms of the macromolecule synthesis rate vR(p,r) (i.e., the growth rate)
vX(p,1−r)=c(r)vR(p,r), |
where c(r):(0,1)→R+ is a positive continuously differentiable function.
As previously described, the reaction vX(p,m) is catalyzed by m, and the reaction vR(p,r) is catalyzed by r, meaning that the ratio between M and R in the microbial culture determines whether the resources are being allocated to the production of biomass or metabolite. This represents the trade-off described in the introduction of this paper, which is here modeled through the function c(r). The fact that c does not depend on the concentration of precursors implies that the host cell has the same affinity to synthesize both biomass and metabolite from the precursors, even when the reactions are not expected to consume the precursors in the same proportion. For the particular case of Michaelis-Menten kinetics, this phenomenon is captured by the half-saturation constant [27]. Notably, for fixed values of r, both synthesis rates are simply proportional. This assumption reduces S1 to
S1:{˙s=D(sin−s)−vM(s,1−r)V,˙p=vM(s,1−r)−μ(p,r)(p+c(r)+1),˙r=(u−r)μ(p,r),˙x=c(r)μ(p,r)V−Dx,˙V=(μ(p,r)−D)V. | (2.8) |
We note that both Assumptions 1 and 2 are formulated for the non-dimensional synthesis rates given in definition (2.6), but they also hold for the original functions vM, vR and vX (taking into account that, for these functions, the domain is defined as R+×[0,1β]→R+).
The asymptotic behavior of system S1 describes the “open-loop” operation mode of the continuous bioreactor, where the resource allocation control u(t) is fixed to ˉu∈(0,1). In the present section we propose a series of mass conservation laws that allow to reduce S1 to a 3-dimensional limiting system. Then, we study the local stability of their equilibria, and we show, using the theory of asymptotically autonomous systems, that the full system S1 converges to the equilibria of its limiting system. Let us start the analysis of system S1 by defining its invariant region in the following lemma.
Lemma 1. The set
Γ={(s,p,r,x,V)∈R5:sin≥s>0,p≥0,x≥0,1≥r≥0,V≥0} |
is positively invariant for the initial value problem.
Proof. Let us analyze the boundaries of Γ,
˙s|s=0=Dsin>0⇒s=0isrepulsive.˙s|s=sin=−vM(s,1−r)V≤0⇒s=sinisrepulsive/invariant.˙p|p=0=vM(s,1−r)≥0⇒p=0isrepulsive/invariant.˙r|r=0=0⇒r=0isinvariant.˙r|r=1=(u−1)μ(p,1)<0⇒r=1isrepulsive.˙x|x=0=vX(p,m)V≥0⇒x=0isrepulsive/invariant.˙V|V=0=0⇒V=0isinvariant. |
We will study the initial value problem of system S1 with initial conditions
sin≥s(0)≥0,p(0)≥0,1≥r(0)>0,x(0)≥0,V(0)>0, | (2.9) |
where two cases have been excluded for being trivial to the analysis: V(0)=0, since it is necessary to have an initial amount of biomass to have bacterial growth; and r(0)=0, since an empty gene expression machinery pool implies null growth rate μ(p,0)=0, and therefore it is not possible to self-replicate from that point.
System S1 can be rewritten as
{˙φ=D(sinvin−vout)+Nvi−vμμ(p,r),˙V=(μ(p,r)−D)V, |
where
● φ≐[s,p,r,m,x]T is the state vector of concentrations in the bioreactor.
● N is the stoichiometry matrix of the macroreactions.
● vi is the vector of internal synthesis rates.
● vin and vout are the vectors of inflows and outflows respectively, associated to the continuous bioreactor setup.
● vμ is the vector modeling the dilution effect due to variation of the bacterial volume.
which are defined as
N≐[−V001−1−10u001−u000V],vi≐[vM(s,1−r)vR(p,r)vX(p,1−r)],vin≐[1,0,0,0,0]T,vout≐diag(φ)[1,0,0,0,1]T,vμ≐diag(φ)[0,1,1,1,0]T. |
By analyzing the left null space of N, it can be seen that there are two mass conservation laws related to the total mass inside the bioreactor.
Definition 1. We define the quantities
w1≐s+(p+m+r)V+x=s+(p+1)V+x, |
w2≐s+(p+rˉu)V+x. |
The first quantity tends asymptotically to w1=sin as t→∞ for every input u(t), as it obeys the dynamical equation
˙w1=D(sin−w1). | (2.10) |
Moreover, when fixing u(t) to ˉu∈(0,1), the quantity w2 also obeys the same Eq (2.10), meaning that this second quantity also converges to sin, which greatly simplifies the analysis of the asymptotic behavior of the system.
Lemma 2. The ω-limit set of any solution of system S1 given by Eq (2.8) lies in the hyperplane
Ω1≐{(s,p,r,x,V)∈R5:s+(p+1)V+x=sin}. |
Moreover, under constant input u(t)=ˉu, this is also true for the hyperplane
Ω2≐{(s,p,r,x,V)∈R5:s+(p+rˉu)V+x=sin}. |
Further on, we will use Lemma 2 to analyze system S1 through its limiting system.
Lemma 2 presents two mass conservation laws that can be used to reduce subsystem S1 by two dimensions. We will first analyze the asymptotic behavior of concentration r: when t→∞, quantities w1=w2=sin, so
s+(p+1)V+x=s+(p+rˉu)V+x⇒r=ˉu |
meaning that, as t→∞, r will converge to the value ˉu. We can also express x=sin−s−(p+1)V, so that the limiting system of S1 becomes
S′1:{˙s=D(sin−s)−ˉvM(s)V,˙p=ˉvM(s)−ˉμ(p)(p+ˉc+1),˙V=(ˉμ(p)−D)V, |
where
ˉvM(s)≐vM(s,1−ˉu),ˉvR(p)≐vR(p,ˉu),ˉvX(p)≐vX(p,1−ˉu),ˉμ(p)≐μ(p,ˉu),ˉc≐c(ˉu). |
Details on the convergence of the limiting system S′1 to the original one S1 will be addressed later in the article. In next section, we will fully describe the asymptotic behavior of S′1.
In the interest of simplifying the notation, we define the following function.
Definition 2. We define the function
ˉf(p)≐ˉvR(p)+ˉvX(p)+ˉμ(p)p=ˉμ(p)(p+ˉc+1), |
Function f meets ˉf(p)>0,ˉf′(p)>0,∀p∈Γ (positive and monotonically increasing).
The main result of local stability study is summarized in Theorem 1.
Theorem 1. The local stability of equilibria is given by the following criterion.
● If ˉμ(pw)≥D:
- The interior equilibrium Ei exists, is unique and locally stable.
- The washout equilibrium Ew exists, is unique and locally unstable.
● If ˉμ(pw)<D:
- The interior equilibrium Ei does not exist.
- The washout equilibrium Ew exists, is unique and locally stable.
In the above criterion, the equilibria are defined as follows:
● The interior equilibrium Ei≐(si,pi,Vi), with
pi:{p∈R+:ˉμ(p)=D}, | (2.11) |
si:{s∈R+:ˉvM(s)=ˉf(pi)}, | (2.12) |
Vi≐D(sin−si)ˉvM(si). | (2.13) |
● The washout equilibrium Ew≐(sin,pw,0), with
pw:{p∈R+:ˉf(p)=ˉvM(sin)}. | (2.14) |
Proof. First, we will prove the boundedness of p. Using definition (2.14), it is possible to define a constant upper-bound on p by analyzing the time-varying upper bound pup(t) with dynamical equation
˙pup≐ˉvM(sin)−ˉf(pup). | (2.15) |
In Eq (2.15), pup converges to the value pw satisfying Eq (2.14). Additionally, the vector field at p=pw is always negative (or null when s=sin)
˙p|p=pw=ˉvM(s)−ˉf(pw)≤0 |
meaning that p=pw is either repulsive or invariant, so a new invariant set Γ′⊂Γ can be defined,
Γ′≐{(s,p,V)∈R3:sin≥s>0,pw≥p>0,V≥0}. | (2.16) |
Equilibrium Ei The existence of the interior equilibrium Ei is given by the boundedness of the flows,
maxˉμ(p)≥D, | (2.17) |
maxˉvM(s)≥ˉf(pi), | (2.18) |
and its uniqueness can be proved through monotonicity arguments: In ˉμ(p)=D, ˉμ(p) is strictly monotonically increasing w.r.t. p so, if inequality (2.17) is met, pi should be unique. Similarly, in ˉvM(s)=ˉf(pi), ˉvM(s) is again strictly monotonically increasing so, if inequality (2.18) is met, si should be unique. As is standard in continuous bioreactors, inequality (2.17) implies that the maximal growth rate of the bacterial population should be bigger than the dilution rate D. At the same time, the inequality (2.18) requires the maximal uptake flow to be bigger than the flows responsible for bacterial growth and metabolite production. In Γ′, these conditions become
ˉμ(pw)≥D, | (2.19) |
where the second condition (2.18) is included in the inequality (2.19), as maxˉvM(s)=ˉvM(sin)=ˉf(pw), and the condition ˉf(pw)≥ˉf(pi) is true if and only if μ(pw)≥D. The Jacobian matrix is given by
Ji=[−D−ˉv′M(s)Vi0−ˉvM(s)ˉv′M(s)ˉf′(p)00ˉμ′(p)Vi0], |
and the characteristic polynomial is
Pi(λ)=(λ+D+ˉv′M(s)Vi)(λ+ˉf′(p))λ+ˉvM(s)ˉv′M(s)ˉμ′(p)Vi | (2.20) |
=λ3+λ2(D+ˉv′M(s)Vi+ˉf′(p))+λ(D+ˉv′M(s)Vi)ˉf′(p)+ˉvM(s)ˉv′M(s)ˉμ′(p)Vi. |
Using Definition 2, it can be seen that −D is an eigenvalue by replacing in expression (2.20)
Pi(−D)=−Dˉv′M(s)Vi(p+ˉc+1)ˉμ′(p)+ˉvM(s)ˉv′M(s)ˉμ′(p)Vi=ˉv′M(s)Viˉμ′(p)˙p=0. |
Then, dividing Pi(λ) by λ+D yields
Pi(λ)1λ+D=λ2+λ(D+ˉv′M(s)Vi+(p+ˉc+1)ˉμ′(p))⏟>0+ˉv′M(s)Vi(p+ˉc+1)ˉμ′(p)⏟>0, |
which, by Routh-Hurwitz criteria, implies that all eigenvalues have negative real part. Then, if it exists, the equilibrium is always stable.
Equilibrium Ew The washout equilibrium Ew exists for all values of Γ′, since the only condition for existence is given by
maxˉf(p)≥ˉvM(sin), |
which is always true in Γ′ since ˉvM(sin)=ˉf(pw) and so the inequality becomes ˉf(pw)≥ˉf(pw). Again, as ˉf(p) is strictly monotonically increasing, there is a unique solution pw. Its stability is given by
Jw=[−D0−ˉcˉv′M(s)−ˉf′(p)000ˉμ(pw)−D] |
with characteristic polynomial
Pw(λ)=(λ+D)(λ+ˉf′(p))(λ−ˉμ(pw)+D). |
Eigenvalues λ=−D and λ=−ˉf′(p) are always real and negative, and so the stability criterion becomes
Ewstableif and only ifˉμ(pw)<D. |
In the current section, we show how the limiting system S′1 can be further reduced using another mass conservation law given by the fact that ˉc is constant. This is formalized in the following lemma.
Lemma 3. The ω-limit set of any solution of the limiting system S′1 lies in the hyperplane
Ω3≐{(s,p,V)∈R3:s+(p+ˉc+1)V=sin} |
Proof. The quantity
w3≐s+(p+ˉc+1)V | (2.21) |
obeys the dynamical equations ˙w3=D(sin−w3), which means that the system converges asymptotically to the limit set Ω3.
Using Lemma 2, we can further reduce system S′1 by expressing
s=sin−V(p+ˉc+1), | (2.22) |
for the values (p,V) that meet sin−V(p+ˉc+1)>0, and so the new limiting system becomes
S″1:{˙p=ˉvM(s(⋅))−ˉμ(p)(p+ˉc+1),˙V=(ˉμ(p)−D)V. |
Since S″1 is a 2-dimensional continuous system, its global behavior (illustrated in Figure 2) can be studied through the Poincaré-Bendixson trichotomy.
Lemma 4. Every solution of the limiting system S″1 with initial conditions (2.9) converges to
● E∗i≐(pi,Vi) if ˉμ(pw)≥D
● E∗w≐(pw,0) if ˉμ(pw)<D
Proof. The Poincaré-Bendixson trichotomy ensures that every non-empty compact ω-limit set of S″1 is either a fixed point, a periodic orbit or a cycle of equilibria. Then, by applying Poincaré-Bendixson theorem through the Dulac criterion, we can discard periodic orbits and cycles of equilibria:
∂∂p˙p+∂∂V˙V=∂∂pˉvM(s(⋅))−ˉμ(p)−ˉμ′(p)(p+ˉc+1)+ˉμ(p)−D |
=ˉv′M(s(⋅))∂s(⋅)∂p−ˉμ′(p)(p+ˉc+1)−D<0, | (2.23) |
as ∂s(⋅)/∂p=−V from Eq (2.22). This ensures that the new limiting system should converge to one of the stable equilibria as t→∞ which, according to Theorem 1, is known to be unique.
Through the theory of asymptotically autonomous systems [22], we can relate the asymptotic behavior of the 2-dimensional limiting system S″1 to that of the full 5-dimensional system S1. The main result is established in Theorem 2.
Theorem 2. Every solution of system S1 with initial conditions (2.9) converges to,
● The extended interior equilibrium ˆEi≐(si,pi,ˉu,xi,Vi) if it exists, with xi≐ˉcVi.
● The extended washout equilibrium ˆEw≐(sin,pw,ˉu,0,0) if ˆEi does not exist.
where the condition for the existence of the extended interior equilibrium is ˉμ(pw)≥D.
Proof. We resort to a more particular case of the general theory of asymptotically autonomous systems (introduced in Appendix F of [24]), that requires a certain number of hypotheses to be met related to the limiting system S″1, its equilibria and the mass conservation equations defined in Lemma 2:
(A1) The dynamical Eqs (2.10) and (2.21) of quantities w1, w2 and w3 are stable.
(A2) S″1 has 2 rest points E∗i and E∗w, which are hyperbolic.
(A3) If E∗i exists, dim(M+(E∗i))=2 and dim(M+(E∗w))=1. If E∗i does not exist, dim(M+(E∗w))=2.
(A4) S″1 has no cycles of rest points, as shown in Eq (2.23).
(A5) S″1 has no periodic orbits, as shown in Eq (2.23).
Then, almost all trajectories of the original system S1 converge to one of the asymptotically stable equilibria of the limiting system, which is always unique.
In this section, we aim to maximize the production of the metabolite of interest X. We start by modeling the synthesis rates as explicit functions of the concentrations of the system. While this latter is not required for most of the results, it enables the numerical simulations both for the dynamical and for the static optimization problems. In particular, we resort to the Michaelis-Menten kinetics defined in previous works [4] and the same biological constants. Then, we pose the problem of maximizing the production of X over a fixed period of time, and we approximate the optimal solution with a constant allocation strategy. This finally motivates the static optimization approach, that is solved in terms of the tuple (D,ˉu).
Michaelis-Menten kinetics are defined as
vM(s,m)≐kMmsKM+s=kM(1−r)sKM+s, |
vR(p,r)≐kRrpKR+p, |
vX(p,m)≐kXmpKX+p=kX(1−r)pKX+p, |
where the biological constants and their values [4,10] are described in Table 1.
Parameter | Description | Unit | Value |
β | Inverse of the cytoplasmic density | Lg−1 | 0.003 |
kM | Rate constant of the metabolic macroreaction | h−1 | 4.32 |
KM | Half-saturation constant of the metabolic macroreaction | gL−1 | 33.33 |
kR | Rate constant of the protein synthesis reaction | h−1 | 3.6 |
KR | Half-saturation constant of the protein synthesis reaction | gL−1 | 1 |
kX | Rate constant of the metabolite synthesis reaction | h−1 | 0.5 |
KX | Half-saturation constant of the metabolite synthesis reaction | gL−1 | 1 |
Assumption 2 is implemented by setting the half-saturation constant KX=KR, such that the function c(r) becomes
c(r)=vX(p,1−r)vR(p,r)=kXkR1−rr. |
The substrate concentration of the inflow sin is set to 0.4 g/L.
The problem of maximizing the production of the metabolite X during a fixed interval of time T is explored in this section. As it is classical in the continuous bioreactor framework, the instantaneous production of metabolite is described by the quantity
DX[gh−1] | (3.1) |
which, using definitions (2.3), can be expressed as DVextx. Then, the total metabolite production over an interval T amounts to
∫T0DVextxdt[g]. | (3.2) |
We will be first interested in dynamically adjusting the agammaegated control u(t) so as to maximize the quantity (3.2) for a fixed dilution rate D. The problem is tackled through an optimal control approach, formulated as
(OCP):{maximizeJx(u)=DVext∫T0x(t)dtsubjecttodynamicsofS1,u(⋅)∈U. |
with U the set of admissible controllers, which are Lebesgue measurable real-valued functions defined on the interval [0,T] and satisfying the constraint u(t)∈[0,1]. We first note that, since D and Vext are constants, the problem reduces to maximizing the integral of the concentration x. Thus, the nature of the OCP allows us to obtain a first result on the existence of the solutions.
Proposition 3. The dynamic maximization problem has at least one solution.
Proof. Since there are no terminal conditions, the set of admissible controls is not empty (any constant control within the prescribed bounds is admissible). For bounded controls (u(t) belongs to [0,1]) and for a fixed final time T>0, the dynamics (2.8) cannot blow up in finite time so all trajectories remain in a fixed compact. Indeed, all state variables but V are bounded (as set (2.16) is invariant), and since
˙V=(μ(p,r)−D)V |
with μ(p,r) bounded, the volume has at worst an exponential rate and is also bounded. As the dynamics is affine in the control, the set of velocities is convex and existence holds by Filippov's theorem [28].
In order to further explore the solution of (OCP), we define the adjoint state vector λ≐(λs,λp,λr,λx,λV) associated to the state vector φ≐(s,p,r,x,V), and we write the Hamiltonian
H(φ,δ,u)=λ0x+⟨λ,F(φ,u)⟩, |
where F represents the right-hand side of system S1 given by Eq (2.8). Developing the expression we obtain
H=λs(D(sin−s)−vM(s,1−r)V)+λp(vM(s,1−r)−vX(p,1−r)−μ(p,r)(p+1))+λr(u−r)μ(p,r)+λx(vX(p,1−r)V−Dx)+λV(μ(p,r)−D)V−λ0x, |
which shows that the Hamiltonian is linear in the control u, so it can be rewritten in the affine form
H=H0+uH1, |
where
H0=λs(D(sin−s)−vM(s,1−r)V)+λp(vM(s,1−r)−vX(p,1−r)−μ(p,r)(p+1)) |
−rλrμ(p,r)+λx(vX(p,1−r)V−Dx)+λV(μ(p,r)−D)V−λ0x,H1=λrμ(p,r). |
In the absence of terminal constraints, there are no abnormal extremals and λ0 can be set to −1. Since the constrained optimal control u should maximize the Hamiltonian, one has
uOCP(t)={0ifH1<0,1ifH1>0,us(t)ifH1=0. |
These alternatives account for the possibility of bang and singular arcs, an extremal being in general an arbitrary concatenation of these. Bang controls correspond to pure allocation strategies u=0 and u=1 (i.e., purely geared towards either the metabolic machinery M or the gene expression machinery R respectively), while singular control us(t) occurs when the function H1 identically vanishes over some subinterval. Such singular arcs can be further described by successively differentiating the switching function H1 until the singular control u can be explicitly computed. For the present case, this occurs when differentiating four times (“order two” singular arc), showing that the singular regime should be entered and leaved through Fuller phenomenon or “chattering” [29] (an infinite number of switchings between the control bounds 0 and 1 over a finite time interval). The theorem hereunder formalizes this statement.
Theorem 4. Any singular arc us(t) of a normal extremal process solution of (OCP) is at least of order two.
Proof. We assume that the process (φ,u,λ) that satisfies PMP is a normal extremal, and we set λ0=−1. In order to further describe the singular arc, we assume H1 vanishes on a whole sub-interval τ=[t1,t2]⊂[0,T]. Then, the switching surface is the set
Σ={(φ,λ)∈R2n|H1=0}, |
where, for this case, n=5. The first derivative is
˙H1=˙λrμ(p,r)+λr(μr(p,r)˙r+μp(p,r)˙p), | (3.3) |
where
μr(p,r)=∂μ(p,r)∂r,μp(p,r)=∂μ(p,r)∂p. |
Along the singular arc H1 is identically zero, so expression (3.3) also vanishes. In order to compute the successive derivatives of H1, we will resort to the Poisson bracket operator [30],
{f,g}=n∑i=1(∂f∂λi∂g∂φi−∂f∂φi∂g∂λi). |
Applying the latter definition to the derivative of H1 we obtain
˙H1=∂H1∂φ˙φ+∂H1∂λ˙λ=n∑i=1(∂H∂λi∂H1∂φi−∂H∂φi∂H1∂λi)={H,H1}={H0+uH1,H1}={H0,H1} |
as {uH1,H1}=u{H1,H1}=0. In this context, we will use the notation H01 to refer to {H0,H1}, and so forth. The second derivative of H1 is
¨H1=˙H01={H,H01}={H0,H01}+u{H1,H01}=H001+uH101. |
Showing that H101=0 along the switching surface Σ ensures that the singular arc is at least of order 2 (otherwise, a singular arc of order 1 could be computed as u=−H001/H101). Since H1 depends only on λr, p and r, the computation of the Poisson bracket reduces to
H101={H1,H01}=∂H1∂λr∂H01∂r−∂H1∂r∂H01∂λr+∂H1∂λp∂H01∂p−∂H1∂p∂H01∂λp=μ(p,r)∂H01∂r−λrμr(p,r)∂H01∂λr−λrμp(p,r)∂H01∂λp | (3.4) |
where
∂H01∂r=˙λrμr(p,r)+λr∂∂r(μr(p,r)˙r+μp(p,r)˙p). | (3.5) |
Replacing the expression (3.5) in Eq (3.4) yields
H101=μ(p,r)[˙λrμr(p,r)+λr∂∂r(μr(p,r)˙r+μp(p,r)˙p)]−λrμr(p,r)∂H01∂λr−λrμp(p,r)∂H01∂λp |
which is also equal to 0 along the switching surface since every term is multiplied either by λr or ˙λr, both identically zero along a singular arc (see H1 expression), showing that the singular arc is at least of order 2.
It is noteworthy that the above proof holds for all general flows vM(s,1−r), vX(p,1−r) and vR(p,r) considered in Assumption 1, with no need to use the defined Michaelis-Menten kinetics. Numerical results shown in Figure 3 confirm that, when the conditions for the existence of the interior equilibrium Ei are met, the optimal control strategy consists on a series of bang-bang arcs, and a singular arc that is entered and left through the chattering phenomenon. While, as already mentioned, it is compulsory to enter and leave order two singular arcs through chattering, a more precise description of the structure of the extremal would require a deeper analysis (for instance to prove that there is only one singular arc). Moreover, when the simulation time is long enough, we observe that the singular control us(t) converges to a constant value, eventually taking the system to steady-state. {This is related to the so-called turnpike phenomenon [31] that relates the singular arc of the dynamic optimization problem with the solution of the static one. See, e.g., [32] for a preliminary analysis on a similar case.
Additionally, we observe that the steady-state approached by the system during the singular regime maximizes the integrand of the cost function, which is in fact the instantaneous production of metabolite described in expression (3.1) (otherwise, it would not be the optimal strategy). Therefore, the static optimal control to which the dynamic one converges along the singular arc is the solution of the optimization problem
(SP):{maximizeJx(ˉu)≐xsubjecttodynamicsofS1,˙φ=0,0≤ˉu≤1. |
The dynamical optimal solution acts as the gold-standard in terms of what can be achieved through control techniques. However, implementing such strategy is, in practice, unfeasible. Consequently, it is possible to design a static suboptimal strategy consisting of a constant allocation usp that takes the system to the same steady-state which maximizes the integrand of the cost function. Numerical simulations of such strategy are shown in Figure 4, where it can be seen that the area below curves xsp and x in OCP only differ slightly (<1% for this particular simulation). The constant control usp basically disregards the initial and final bang arcs, as well as the chattering phenomena, which constitutes a small fraction of the complete time horizon T. Moreover, this fraction gets smaller as T becomes large, which is a typical feature of the turnpike effect for a specific class of optimal control problems [31]. This way, the difference between strategies becomes marginal in long-term production schemes.
Additionally, we provide a numerical computation of the successive derivatives of the switching function H1 in Figure 5. It can be seen that the fourth derivative H10001≠0 over the interval where H1=0, which shows that the singular arc is of order 2. Furthermore, we verify that the Kelley (or generalized Legendre-Clebsch [34]) condition
(−1)k∂∂u[d2kdt2k(∂H∂u)]<0,∀t∈I | (3.6) |
is met along the singular arc, which in this case is equivalent to H10001<0,∀t∈I. A check of this condition, necessary for optimality, is also shown in Figure 5. Although there is no available sufficient condition to test local optimality of extremals with Fuller arcs, verifying the Legendre-Clebsch condition along the singular arc only ensures that we do not compute a too crude local minimizer. Besides, numerically only a small finite number of bang arcs are retrieved by the optimizer for the chattering parts before and after the singular arc. This is usually sufficient to give a very good approximation of the solution.
We have shown that a constant allocation decision ˉu represents a simplified alternative to the optimal control solution, a strategy composed of bang arcs, a time-varying singular arc and the chattering artifact. In this section, we will further explore the static optimization problem by adding a second degree of freedom to the problem: the dilution rate D. In addition to that, we investigate two objectives: the production of biomass V and of metabolite X. The static biomass maximization problem (SPV) can be written as
(SPV):{maximizeJV(ˉu,D)≐DVsubjecttodynamicsofS1,˙φ=0,0≤ˉu≤1. |
Analogously, the product maximization problem can be defined as
(SPX):{maximizeJX(ˉu,D)≐DX=DVextxsubjecttodynamicsofS1,˙φ=0,0≤ˉu≤1. |
Since we look for the steady-states that maximize each objective, the washout equilibrium Ew can be excluded from the analysis since, as shown in Theorem 2, the equilibrium corresponds to the steady-state values V=0 and x=0. Therefore, the static problems are reduced to finding the equilibria Ei in terms of the pair (D,ˉu) that maximize each objective function. Moreover, it can be shown that the optimal solution cannot belong to the boundary of the equilibrium Ei
Θ≐{(ˉu,D)∈R2:ˉμ(pw)=D}. |
Indeed, using the definitions (2.11), (2.12) and (2.13), it can be seen that on Θ there is no bacterial population, as si=sin, which means that Vi=xi=0. We formalize the latter reasoning in the following proposition.
Proposition 5. Every solution (D,ˉu) of the static optimization problems (SPV) and (SPX) is in the region of existence of Ei given by the condition ˉμ(pw)>D.
When considering the same self-replicator scheme under constant environmental conditions (which could describe fed-batch cultivation) the solution for both static problems corresponds to the steady-state with maximal growth rate [10]. Proposition 5 shows that this is not the case in continuous bioreactors, where maximal growth rate is attained at the boundary set of existence of the equilibrium Ei. Two interesting particular cases within Θ are the pure static allocation strategies ˉu=0 and ˉu=1: A pure metabolic strategy ˉu=0 will lead to a bacterial population with no RNA polymerase (i.e., r=0), which will eventually stop the production of biomass (as vR(p,0)=0), leading to washout in the bioreactor. Analogously, allocating all resources to the gene expression machinery will finally empty the metabolic machinery m, halting the absorption of substrate from the environment (as vM(p,0)=0) and depleting the bacteria from precursors, which will also lead to washout. From this analysis, we can conclude that any optimal steady-state allocation ˉu should belong to (0,1).
We recall that, in continuous bioreactors, the growth rate ˉμ(pi) is fixed by the dilution rate D, as shown in Eq (2.11). Figure 6 illustrates a numerical analysis of the static problems. It is interesting to notice in both subfigures 6a and 6b that the model accounts for the classical quasi-linear relation between the maximal growth rate (that lies on the boundary Θ), and the control ˉu which regulates the RNA/protein mass ratio of the bacterial population [35]. The latter is a phenomenon which has been first observed experimentally, and later used to develop dynamical self-replicator models for natural and biotechnological purposes [4,10,11]. However, in the present case, the growth rate is fixed through D, so it is not a result of the nutrient quality in the environment, which enables the multivariate approach.
Biomass maximization objective Results for this problem are shown in Figure 6a. For this case, the curve ˉuopt remains over 0.6 for all values of the growth rate, tending to ˉuopt=1 as the growth rate goes to 0. This shows that, in order to maximize V, the allocation strategy should prioritize the synthesis of macromolecules of the gene expression machinery R over the metabolic machinery M, which catalyzes the production of biomass and not the synthesis of metabolites. Moreover, the maximum DVi is attained through a fairly high growth rate (≈85% of the maximal growth rate).
Product maximization objective Results for this case are shown in Figure 6b. We can see that, in opposition to the first case, the optimal solution requires allocating as much precursors as possible into the metabolic machinery M, no matter the value of the dilution rate. In other words, the allocation control ˉu should be as low as possible within the region of existence of the equilibrium Ei (but not in Θ). This result is consistent with the fact that the metabolic machinery M catalyzes the synthesis of metabolite X. It is noteworthy that the optimal point DXi is accomplished through a rather lower dilution rate D in comparison to the biomass production case, resulting in a continuous production at a rate of about 35% of the maximal growth rate. The latter might appear counter-intuitive, as it is well established in the literature that high dilution rates in continuous bioreactors imply high production rates. This characteristic can be attributed to the compromise between allocating resources to the metabolic machinery M and increasing the dilution rate, linked to the maximum value of the function c(r). In other words, the lower the dilution rate of operation, the wider the interval of existence of the equilibrium (umin,umax), which enables the possibility of further promoting the synthesis of compounds of the metabolic machinery (by artificially lowering the allocation parameter ˉu) without going to washout. Figure 7a illustrates the resulting synthesis rates for each solution. In (SPX), the synthesis rate of the metabolite vX is increased at the expense of reducing both precursor and biomass synthesis rates vM and vR, which is in large part due to the reduction of the dilution rate D. This difference in the flux distribution impacts directly on the mass quantities inside the bioreactor at steady-state—depicted in Figure 7b—for each problem: for the (SPX) case, there is a reduction in the biomass V of only 20% w.r.t. the (SPV) case. However, we can see an increase on the amount of metabolite X of about 5 times that of the (SPV) case. As already said, the allocation parameter ˉu becomes quite lower in the metabolite synthesis objective. In turn, this inhibits the synthesis of macromolecules, which translates into a decrease of the bacterial population's growth rate. To compensate this effect, the steady-state pool of precursors P required for producing the metabolite X becomes considerably bigger than that of the biomass production objective (about 10 times), therefore increasing the rate of fabrication of biomass vR(p,r).
The objective in this paper was to synthesize a certain metabolite of interest by re-adjusting the natural allocation of resources in bacteria. This is done by drawing resources from the native pathways of the cell originally used for producing biomass, and allocating them into the production of the compound X. We addressed the problem in the continuous bioreactor framework, which allows a steady-state production regime. Based on previous dynamical systems approaches [4,10,12], we proposed a coarse-grained self-replicator model capable of accounting for well-studied bacterial growth laws in a simplified way, and we studied its asymptotic behavior. We tackled two different production objectives: Biomass maximization V and metabolite maximization X, and we compared results in order to better understand the potential control strategies required to that effect. We rely on novel bio-engineering techniques capable of delivering groundbreaking control schemes: A synthetic growth switch that allows to control the transcription of RNA polymerase through an inducible promoter. In addition to this regulation mechanism, we include the dilution rate as a control input, which yields a multi-variable optimization problem. We concluded by showing very contrasting results, but in accordance with our previous understanding of microbial resource allocation. The biomass-oriented strategy involves an almost maximal dilution rate, and prioritizes investing resources into the gene expression machinery. Conversely, producing the compound of interest requires a rather low value of the dilution rate, which allows an allocation strategy more geared towards the synthesis of components of the metabolic machinery (i.e., a lower value of ˉu). The latter shows there exist a compromise between augmenting the dilution rate and artificially boosting the production of enzymes and transporters.
The capacity to interfere with the natural bacterial behaviors at molecular level represents a promising tool to improve biotechnological processes. We are interested in further exploring the details of the implementation of such control schemes, by considering different models of the external growth switch. A next step in this regard would involve taking into account the nature of the external signal I and its physical constraints, in order to adapt our current results towards a more implementable control loop. Additionally, our approach is based on very simplified representations of bacterial growth. Such biological processes usually involve numerous reactions and variables. In our case, we clustered all macromolecules into only 2 different classes, and we purposely omitted a number of known phenomena in bacteria, such as cell division, protein degradation and the influence of temperature. However, some of these effects have been proven to affect only marginally the results regarding the allocation problem [36]. Thus, in our case, a simplified representation becomes useful to emphasize the effects of optimally dealing with the internal resource distribution of bacteria in industrial frameworks. Eventually, such optimal strategies could provide guidance for developing online feedback solutions based on real-time measuring of the process.
This work was partially supported by ANR project Maximic (ANR-17-CE40-0024-01), Inria IPL Cosy and Labex SIGNALIFE (ANR-11-LABX-0028-01). We acknowledge the support of the FMJH Program PGMO and the support to this program from EDF-THALES-ORANGE. We also thank Hidde de Jong (INRIA Grenoble - Rhône-Alpes) for discussions on the manuscript.
All authors declare no conflicts of interest in this paper.
[1] |
Abbad MM (2013) E-banking in Jordan. Behav Inform Technol 32: 681-694. doi: 10.1080/0144929X.2011.586725
![]() |
[2] | Aboelmaged MGG, Gebba TR (2013) Mobile banking adoption: an examination of technology acceptance model and theory of planned behavior. Int J Bus Res Dev 2: 35-50. |
[3] |
Ajzen I (1991) The theory of planned behaviour. Organ Behav Hum Decis Process 50: 179-211. doi: 10.1016/0749-5978(91)90020-T
![]() |
[4] |
Ajzen I (2011) The theory of planned behaviour: reactions and reflections. Psychol Health 26: 1113-1127. doi: 10.1080/08870446.2011.613995
![]() |
[5] | Alaeddin O, Altounjy R (2018) Trust, technology awareness and satisfaction effect into the intention to use cryptocurrency among generation Z in Malaysia. Int J Eng Technol 7: 8-10. |
[6] |
Aldás-Manzano J, Lassala-Navarre C, Ruiz-Mafe C, et al. (2009) The role of consumer innovativeness and perceived risk in online banking usage. Int J Bank Mark 27: 53-75. doi: 10.1108/02652320910928245
![]() |
[7] | Anderson JC, Gerbing DW (1988) Structural equation modeling in practice: A review and recommended two-step approach. Psychol Bull 103: 411. |
[8] | Azjen I, Fishbein M (1980) Understanding attitudes and predicting social behavior, Englewood Cliffs, NJ: Prentice-Hall. |
[9] |
Bailey AA, Pentina I, Mishra AS, et al. (2017) Mobile payments adoption by US consumers: an extended TAM. Int J Retail Distrib Manage 45: 626-640. doi: 10.1108/IJRDM-08-2016-0144
![]() |
[10] | Bansal S, Bansal A, Blake MB (2010) Trust-based dynamic web service composition using social network analysis, 2010 IEEE International Workshop on: Business Applications of Social Network Analysis (BASNA), 1-8. |
[11] |
Bentler PM, Bonett DG (1980) Significance tests and goodness of fit in the analysis of covariance structures. Psychol Bull 88: 588-606. doi: 10.1037/0033-2909.88.3.588
![]() |
[12] | Carmines EG, McIver JP (1983) An introduction to the analysis of models with unobserved variables. Political Methodol 9: 51-102. |
[13] | Cham TH, Low SC, Lim CS, et al. (2018) Preliminary study on consumer attitude towards fintech products and services in Malaysia. Int J Eng Technol 7: 166-169. |
[14] | Chang YP, Lan LY, Zhu DH (2018) Understanding the intention to continue use a mobile payment. Int J Bus Inform 12: 363-390. |
[15] |
Chau PY, Hu PJ (2002) Examining a model of information technology acceptance by individual professionals: an exploratory study. J Manage Inform Syst 18: 191-229. doi: 10.1080/07421222.2002.11045699
![]() |
[16] |
Chen JF, Chang JF, Kao CW, et al. (2016) Integrating ISSM into TAM to enhance digital library services. Electron Lib 34: 58-73. doi: 10.1108/EL-01-2014-0016
![]() |
[17] |
Cheng TE, Lam DY, Yeung AC (2006) Adoption of internet banking: an empirical study in Hong Kong. Decis Support Syst 42: 1558-1572. doi: 10.1016/j.dss.2006.01.002
![]() |
[18] |
Chong A, Ooi K, Lin B, et al. (2010) Online banking adoption: an empirical analysis. Int J Bank Mark 28: 267-287. doi: 10.1108/02652321011054963
![]() |
[19] |
Chua CJ, Lim CS, Aye AK (2019) Factors affecting the consumer acceptance towards fintech products and services in Malaysia. Int J Asian Soc Sci 9: 59-65. doi: 10.18488/journal.1.2019.91.59.65
![]() |
[20] | Chuang LM, Liu CC, Kao HK (2016) The adoption of fintech service: TAM perspective. Int J Manage Admin Sci 3: 1-15. |
[21] | Chuen DLK (2015) Handbook of Digital Currency: Bitcoin, Innovation, Financial Instruments, and Big Data, Academic Press. |
[22] |
Datta PA (2011) Preliminary study of ecommerce adoption in developing countries. Inform Syst J 21: 3-32. doi: 10.1111/j.1365-2575.2009.00344.x
![]() |
[23] |
Davis FD, Bagozzi RP, Warshaw PR (1989) User acceptance of computer technology: a comparison of two theoretical models. Manage Sci 35: 982-1003. doi: 10.1287/mnsc.35.8.982
![]() |
[24] |
Demoulin NTM, Djelassi S (2016) An integrated model of self-service technology (SST) usage in a retail context. Int J Retail Distrib Manage 44: 540-559. doi: 10.1108/IJRDM-08-2015-0122
![]() |
[25] |
Devlin JF (1997) Adding value to retail financial services. J Mark Pract 3: 251-267. doi: 10.1108/EUM0000000004461
![]() |
[26] |
Donner J, Tellez CA (2008) Mobile banking and economic development: Linking adoption, impact, and use. Asian J Commun 18: 318-332. doi: 10.1080/01292980802344190
![]() |
[27] |
Dulcic Z, Pavlic D, Silic I (2012) Evaluating the intended use of decision support system (DSS) by applying technology acceptance model (TAM) in business organizations in Croatia. Procedia-Soc Behav Sci 58: 1565-1575. doi: 10.1016/j.sbspro.2012.09.1143
![]() |
[28] | 2020 Fintech Vietnam Report and Startup Map: Fintech startups Tripled since 2017, 2020. Fintech Singapore, 2020. Available from: https://fintechnews.sg/45354/vietnam/2020-fintech-vietnam-report-and-startup-map/. |
[29] | Fishbein M, Ajzen I (1977) Belief, attitude, intention, and behavior: an introduction to theory and research. Contemp Sociol 6 : 244. |
[30] |
Forsythe SM, Shi B (2003) Consumer patronage and risk perceptions in Internet shopping. J Bus Res 56: 867-875. doi: 10.1016/S0148-2963(01)00273-9
![]() |
[31] | FSB (2019) Financial Stability Implications from Fintech. Financial Stability Board. |
[32] | Global Innovation Index 2020. World Intellectual Property Organization, 2020. Available from: https://www.wipo.int/global_innovation_index/en/2020/. |
[33] |
Grabner-Kräuter S, Faullant R (2008) Consumer acceptance of Internet banking: the influence of internet trust. Int J Bank Mark 26: 483-504. doi: 10.1108/02652320810913855
![]() |
[34] | Group VMW (2018) Report "Application of Financial Technology (Fintech) in Microfinance activities towards Universal Finance in Vietnam", Vietnam. |
[35] |
Gupta A, Arora N (2017) Consumer adoption of m-banking: a behavioral reasoning theory perspective. Int J Bank Mark 35: 733-747. doi: 10.1108/IJBM-11-2016-0162
![]() |
[36] | Ha H (2004) Factors influencing consumer perceptions of brand trust online J Prod Brand Manage 13: 329-342. |
[37] | Hair JF (1998) Multivariate data analysis, Upper Saddle River, N.J., Prentice Hall. |
[38] |
Halilovic S, Cicic M (2013) Antecedents of information systems user behavior-extended expectation-confirmation model. Behav Inform Technol 32: 359-370. doi: 10.1080/0144929X.2011.554575
![]() |
[39] |
Hamidi H, Safareeyeh M (2019) A model to analyze the effect of mobile banking adoption on customer interaction and satisfaction: a case study of m-banking in Iran. Telematics Inform 38: 166-181. doi: 10.1016/j.tele.2018.09.008
![]() |
[40] |
Hanafizadeh P, Behboudi M, Koshksaray AA, et al. (2014) Mobile-banking adoption by Iranian bank clients. Telematics Inform 31: 62-78. doi: 10.1016/j.tele.2012.11.001
![]() |
[41] | Hu Z, Ding S, Li S, et al. (2019) Adoption intention of fintech services for bank users: An empirical examination with an extended technology acceptance model. Symmetry 11: 340. |
[42] |
Huh HJ, Kim TT, Law R (2009) A comparison of competing theoretical models for understanding acceptance behavior of information systems in upscale hotels. Int J Hosp Manage 28: 121-134. doi: 10.1016/j.ijhm.2008.06.004
![]() |
[43] | Jiwasiddi A, Adhikara C, Adam M, et al. (2019) Attitude toward using Fintech among Millennials. The 1st Workshop on Multimedia Education, Learning, Assessment and its Implementation in Game and Gamification in conjunction with COMDEV 2018, European Alliance for Innovation (EAI). |
[44] | Jonker N (2019) What drives the adoption of crypto-payments by online retailers? Electron Comm Res Appl 35: 100848. |
[45] |
Kapoor K, Dwivedi Y, Piercy NC, et al. (2014) RFID integrated systems in libraries: extending TAM model for empirically examining the use. J Enterp Inform Manage 27: 731-758. doi: 10.1108/JEIM-10-2013-0079
![]() |
[46] |
Kesharwani A, Singh Bisht S (2012) The impact of trust and perceived risk on internet banking adoption in India: An extension of technology acceptance model. Int J Bank Mark 30: 303-322. doi: 10.1108/02652321211236923
![]() |
[47] |
Khedmatgozar HR, Shahnazi A (2018) The role of dimensions of perceived risk in adoption of corporate internet banking by customers in Iran. Electron Comm Res 18: 389-412. doi: 10.1007/s10660-017-9253-z
![]() |
[48] | Khodabandeh A, Lindh C (2020) The importance of brands, commitment, and influencers on purchase intent in the context of online relationships. Australas Mark J. |
[49] |
Kim S (2012) Factors affecting the use of social software: TAM perspectives. Electron Lib 30: 690-706. doi: 10.1108/02640471211275729
![]() |
[50] |
Kim YG, Woo E (2016) Consumer acceptance of a quick response (QR) code for the food traceability system: application of an extended technology acceptance model (TAM). Food Res Int 85: 266-272. doi: 10.1016/j.foodres.2016.05.002
![]() |
[51] |
Kim YJ, Park YJ, Choi J, et al. (2015) An empirical study on the adoption of "fintech" service: focused on mobile payment services. Adv Sci Technol Lett 114: 136-140. doi: 10.14257/astl.2015.114.26
![]() |
[52] |
Koksal MH (2016) The intentions of Lebanese consumers to adopt mobile banking. Int J Bank Mark 34: 327-346. doi: 10.1108/IJBM-03-2015-0025
![]() |
[53] |
Koufaris M (2002) Applying the technology acceptance model and flow theory to online consumer behavior. Inform Syst Res 13: 205-223. doi: 10.1287/isre.13.2.205.83
![]() |
[54] |
Lee DY, Lehto MR (2013) User acceptance of YouTube for procedural learning: an extension of the technology acceptance model. Comput Educ 61: 193-208. doi: 10.1016/j.compedu.2012.10.001
![]() |
[55] |
Lee MKO, Turban E (2001) A trust model for consumer internet shopping. Int J Electron Comm 6: 75-91. doi: 10.1080/10864415.2001.11044227
![]() |
[56] |
Leicht T, Chtourou A, Youssef KB (2018) Consumer innovativeness and intentioned autonomous car adoption. J High Technol Manage Res 29: 1-11. doi: 10.1016/j.hitech.2018.04.001
![]() |
[57] |
Liebermann Y, Stashevsky S (2002) Perceived risks as barriers to Internet and e-commerce usage. Qual Mark Res 5: 291-300. doi: 10.1108/13522750210443245
![]() |
[58] |
Lien NTK, Thi Doan TT, Bui TN (2020) Fintech and banking: evidence from Vietnam. J Asian Financ Econ Bus 7: 419-426. doi: 10.13106/jafeb.2020.vol7.no9.419
![]() |
[59] |
Lin WR, Wang YH, Hung YM (2020) Analyzing the factors influencing adoption intention of internet banking: Applying DEMATEL-ANP-SEM approach. PlOS ONE 15: e0227852. doi: 10.1371/journal.pone.0227852
![]() |
[60] |
López-Nicolás C, Molina-Castillo FJ, Bouwman H (2008) An assessment of advanced mobile services acceptance: Contributions from TAM and diffusion theory models. Inform Manage 45: 359-364. doi: 10.1016/j.im.2008.05.001
![]() |
[61] |
Lockett A, Littler D (1997) The adoption of direct banking services. J Mark Manage 13: 791-811. doi: 10.1080/0267257X.1997.9964512
![]() |
[62] |
Malaquias RF, Hwang Y (2016) An empirical study on trust in mobile banking: A developing country perspective. Comput Hum Behav 54: 453-461. doi: 10.1016/j.chb.2015.08.039
![]() |
[63] |
Mallat N, Rossi M, Tuunainen VK (2009) The impact of use context on mobile services acceptance: The case of mobile ticketing. Inform Manage 46: 190-195. doi: 10.1016/j.im.2008.11.008
![]() |
[64] |
Manrai LA, Manrai AK (2007) A field study of customers' switching behavior for bank services. J Retailing Consum Serv 14: 208-215. doi: 10.1016/j.jretconser.2006.09.005
![]() |
[65] |
Marakarkandy B, Yajnik N, Dasgupta C (2017) Enabling internet banking adoption: an empirical examination with an augmented technology acceptance model (Tam). J Enterp Inform Manage 30: 263-294. doi: 10.1108/JEIM-10-2015-0094
![]() |
[66] |
Martins C, Oliveira T, Popovič A (2014) Understanding the Internet banking adoption: A unified theory of acceptance and use of technology and perceived risk application. Int J Inform Manage 34: 1-13. doi: 10.1016/j.ijinfomgt.2013.06.002
![]() |
[67] |
McKnight DH, Chervany NL (2001) What trust means in e-commerce customer relationships: An interdisciplinary conceptual typology. Int J Electron Comm 6: 35-59. doi: 10.1080/10864415.2001.11044235
![]() |
[68] |
Meyliana M, Fernando E, Surjandy S (2019) The influence of perceived risk and trust in adoption of fintech services in Indonesia. CommIT J 13: 31-37. doi: 10.21512/commit.v13i1.5708
![]() |
[69] | Miller K (2005) Communication theories, USA: Macgraw-Hill. |
[70] |
Moon JW, Kim YG (2001) Extending the TAM for a World-Wide-Web context. Inform Manage 38: 217-230. doi: 10.1016/S0378-7206(00)00061-6
![]() |
[71] | Munoz-Leiva F, Climent-Climent S, Liébana-Cabanillas F (2017) Determinants of intention to use the mobile banking apps: An extension of the classic TAM model. Span J Mark-ESIC 21: 25-38. |
[72] |
Ng AW, Kwok BKB (2017) Emergence of Fintech and cybersecurity in a global financial centre: Strategic approach by a regulator. J Financ Regul Compliance 25: 422-434. doi: 10.1108/JFRC-01-2017-0013
![]() |
[73] |
Olson JM, Zanna MP (1993) Attitudes and attitude change. Annu Rev Psycho 44: 117-154. doi: 10.1146/annurev.ps.44.020193.001001
![]() |
[74] |
Patel KJ, Patel HJ (2018) Adoption of internet banking services in Gujarat: An extension of TAM with perceived security and social influence. Int J Bank Mark 36: 147-169. doi: 10.1108/IJBM-08-2016-0104
![]() |
[75] | Pauline R (2003) Inter-Organizational-Trust in Business to Business E-Commerce: A Case Study in Customs Clearance. J Glob Inform Manage 11: 1-19. |
[76] |
Phan DTT, Nguyen TTH, Bui TA (2019) Going beyond border? Intention to use international bank cards in Vietnam. J Asian Financ Econ Bus 6: 315-325. doi: 10.13106/jafeb.2019.vol6.no3.315
![]() |
[77] |
Qiu LY, Li D (2008) Applying TAM in B2C E-commerce research: An extended model. Tsinghua Sci Technol 13: 265-272. doi: 10.1016/S1007-0214(08)70043-9
![]() |
[78] | Ratnasingam P (2003) Inter-organizational-trust in business to business e-commerce: A case study in customs clearance. J Glob Inform Manage 11: 1-19. |
[79] | Ratten V (2014) A US-China comparative study of cloud computing adoption behavior: The role of consumer innovativeness, performance expectations and social influence. J Entrepreneurship Emerg Econ 6: 53-71. |
[80] |
Riquelme EH, Rios RE (2010) The moderating effect of gender in the adoption of mobile banking. Int J Bank Mark 28: 328-341. doi: 10.1108/02652321011064872
![]() |
[81] | Romānova I, Kudinska M (2016) Banking and fintech: A challenge or opportunity? Contemp Stud Econ Financ Anal 98: 21-35. |
[82] |
Ruparelia N, White L, Hughes K (2010) Drivers of brand trust in internet retailing. J Prod Brand Manage 19: 250-260. doi: 10.1108/10610421011059577
![]() |
[83] |
Sánchez-Torres JA, Canada FJA, Sandoval AV, et al. (2018) E-banking in Colombia: factors favouring its acceptance, online trust and government support. Int J Bank Mark 36: 170-183. doi: 10.1108/IJBM-10-2016-0145
![]() |
[84] |
Shaikh IM, Qureshi MA, Noordin K, et al. (2020) Acceptance of Islamic financial technology (FinTech) banking services by Malaysian users: an extension of technology acceptance model. Foresight 22: 367-383. doi: 10.1108/FS-12-2019-0105
![]() |
[85] | Smith AA, Synowka DP, Smith AD (2014) E-commerce quality and adoptive elements of e-ticketing for entertainment and sporting events. Int J Bus Inform Syst 15: 450-487. |
[86] |
Steiger JH (1990) Structural model evaluation and modification: an interval estimation approach. Multivar Behav Res 25: 173-180. doi: 10.1207/s15327906mbr2502_4
![]() |
[87] |
Stewart H, Jürjens J (2018) Data security and consumer trust in FinTech innovation in Germany. Inform Comput Secur 26: 109-128. doi: 10.1108/ICS-06-2017-0039
![]() |
[88] |
Stocchi L, Michaelidou N, Micevski M (2019) Drivers and outcomes of branded mobile app usage intention. J Prod Brand Manage 28: 28-49. doi: 10.1108/JPBM-02-2017-1436
![]() |
[89] |
Teo T, Lee BC, Chai CS, et al. (2009) Assessing the intention to use technology among pre-service teachers in Singapore and Malaysia: A multigroup invariance analysis of the Technology Acceptance Model (TAM). Comput Educ 53: 1000-1009. doi: 10.1016/j.compedu.2009.05.017
![]() |
[90] |
Thakur R, Srivastava M (2015) A study on the impact of consumer risk perception and innovativeness on online shopping in India. Int J Retail Distrib Manage 43: 148-166. doi: 10.1108/IJRDM-06-2013-0128
![]() |
[91] |
Venkatesh V, Davis FD (2000) A theoretical extension of the technology acceptance model: Four longitudinal field studies. Manage Sci 46: 186-204. doi: 10.1287/mnsc.46.2.186.11926
![]() |
[92] |
Venkatesh V, Morris MG, Davis GB, et al. (2003) User acceptance of information technology: toward a unified view. MIS Quart 27: 425-478. doi: 10.2307/30036540
![]() |
[93] | Vietnam Fintech Report 2020. Fintech Singapore, 2020. Available from: https://fintechnews.sg/wp-content/uploads/2020/11/Vietnam-Fintech-Report-2020.pdf. |
[94] | Vietnam Microfinance Working Group. Annual Report, 2018. Available from: https://microfinance.vn/wp-content/uploads/2019/06/gg.pdf. |
[95] |
Wallace LG, Sheetz SD (2014) The adoption of software measures: A technology acceptance model (TAM) perspective. Inform Manage 51: 249-259. doi: 10.1016/j.im.2013.12.003
![]() |
[96] |
Wang WT, Li HM (2012) Factors influencing mobile services adoption: A brand-equity perspective. Internet Res 22: 142-179. doi: 10.1108/10662241211214548
![]() |
[97] |
Wang Y, Wang Y, Lin H, et al. (2003) Determinants of user acceptance of Internet banking: An empirical study. Int J Serv Ind Manage 14: 501-519. doi: 10.1108/09564230310500192
![]() |
[98] |
Wonglimpiyarat J (2017) FinTech banking industry: a systemic approach. Foresight 19: 590-603. doi: 10.1108/FS-07-2017-0026
![]() |
[99] |
Wu JH, Wang SC (2005) What drives mobile commerce?: An empirical evaluation of the revised technology acceptance model. Inform Manage 42: 719-729. doi: 10.1016/j.im.2004.07.001
![]() |
[100] |
Yang HD, Yoo Y (2004) It's all about attitude: revisiting the technology acceptance model. Decis Support Syst 38: 19-31. doi: 10.1016/S0167-9236(03)00062-9
![]() |
[101] | Zhang T, Lu C, Kizildag M (2018) Banking "on-the-go": examining consumers' adoption of mobile banking services. Int J Qual Serv Sci 10: 279-295. |
[102] | Zhou L, Yang D (2011) An enterprise mobile e-business system model based on pushmail, 2011 IEEE 2nd International Conference on Software Engineering and Service Science, 730-733. |
![]() |
![]() |
1. | Zhihua Chen, Xinke Wang, Peng Gao, Hongju Liu, Bosheng Song, Predicting Disease Related microRNA Based on Similarity and Topology, 2019, 8, 2073-4409, 1405, 10.3390/cells8111405 | |
2. | Lei Zheng, Dongyang Liu, Wuritu Yang, Lei Yang, Yongchun Zuo, RaacLogo: a new sequence logo generator by using reduced amino acid clusters, 2020, 1467-5463, 10.1093/bib/bbaa096 | |
3. | Leyi Wei, Wenjia He, Adeel Malik, Ran Su, Lizhen Cui, Balachandran Manavalan, Computational prediction and interpretation of cell-specific replication origin sites from multiple eukaryotes by exploiting stacking framework, 2020, 1467-5463, 10.1093/bib/bbaa275 | |
4. | Chaolu Meng, Shunshan Jin, Lei Wang, Fei Guo, Quan Zou, AOPs-SVM: A Sequence-Based Classifier of Antioxidant Proteins Using a Support Vector Machine, 2019, 7, 2296-4185, 10.3389/fbioe.2019.00224 | |
5. | Xing Gao, Guilin Li, A Cytokine Protein Identification Model Based on the Compressed PseKRAAC Features, 2020, 8, 2169-3536, 141422, 10.1109/ACCESS.2020.3013409 | |
6. | Meijie Zhang, Luyang Cheng, Yina Zhang, Characterization of Dysregulated lncRNA-Associated ceRNA Network Reveals Novel lncRNAs With ceRNA Activity as Epigenetic Diagnostic Biomarkers for Osteoporosis Risk, 2020, 8, 2296-634X, 10.3389/fcell.2020.00184 | |
7. | Yu-Miao Chen, Xin-Ping Zu, Dan Li, Identification of Proteins of Tobacco Mosaic Virus by Using a Method of Feature Extraction, 2020, 11, 1664-8021, 10.3389/fgene.2020.569100 | |
8. | Qingwen Li, Wenyang Zhou, Donghua Wang, Sui Wang, Qingyuan Li, Prediction of Anticancer Peptides Using a Low-Dimensional Feature Model, 2020, 8, 2296-4185, 10.3389/fbioe.2020.00892 | |
9. | Xudong Zhao, Hanxu Wang, Hangyu Li, Yiming Wu, Guohua Wang, Identifying Plant Pentatricopeptide Repeat Proteins Using a Variable Selection Method, 2021, 12, 1664-462X, 10.3389/fpls.2021.506681 | |
10. | Hasan Zulfiqar, Muhammad Shareef Masoud, Hui Yang, Shu-Guang Han, Cheng-Yan Wu, Hao Lin, Balachandran Manavalan, Screening of Prospective Plant Compounds as H1R and CL1R Inhibitors and Its Antiallergic Efficacy through Molecular Docking Approach, 2021, 2021, 1748-6718, 1, 10.1155/2021/6683407 | |
11. | Wei Chen, Kewei Liu, Analysis and Comparison of RNA Pseudouridine Site Prediction Tools, 2020, 15, 15748936, 279, 10.2174/1574893614666191018171521 | |
12. | Lei Xu, Guangmin Liang, Baowen Chen, Xu Tan, Huaikun Xiang, Changrui Liao, A Computational Method for the Identification of Endolysins and Autolysins, 2020, 27, 09298665, 329, 10.2174/0929866526666191002104735 | |
13. | Fang Wang, Zheng-Xing Guan, Fu-Ying Dao, Hui Ding, A Brief Review of the Computational Identification of Antifreeze Protein, 2019, 23, 13852728, 1671, 10.2174/1385272823666190718145613 | |
14. | Zhiyu Tao, Benzhi Dong, Zhixia Teng, Yuming Zhao, The Classification of Enzymes by Deep Learning, 2020, 8, 2169-3536, 89802, 10.1109/ACCESS.2020.2992468 | |
15. | Meng-Lu Liu, Wei Su, Jia-Shu Wang, Yu-He Yang, Hui Yang, Hao Lin, Predicting Preference of Transcription Factors for Methylated DNA Using Sequence Information, 2020, 22, 21622531, 1043, 10.1016/j.omtn.2020.07.035 | |
16. | Zi-Mei Zhang, Zheng-Xing Guan, Fang Wang, Dan Zhang, Hui Ding, Application of Machine Learning Methods in Predicting Nuclear Receptors and their Families, 2020, 16, 15734064, 594, 10.2174/1573406415666191004125551 | |
17. | Kuan Li, Yue Zhong, Xuan Lin, Zhe Quan, Predicting the Disease Risk of Protein Mutation Sequences With Pre-training Model, 2020, 11, 1664-8021, 10.3389/fgene.2020.605620 | |
18. | Chen-Chen Li, Bin Liu, MotifCNN-fold: protein fold recognition based on fold-specific features extracted by motif-based convolutional neural networks, 2020, 21, 1467-5463, 2133, 10.1093/bib/bbz133 | |
19. | Lifu Zhang, Benzhi Dong, Zhixia Teng, Ying Zhang, Liran Juan, Identification of Human Enzymes Using Amino Acid Composition and the Composition of k-Spaced Amino Acid Pairs, 2020, 2020, 2314-6133, 1, 10.1155/2020/9235920 | |
20. | Chunyu Wang, Kai Sun, Juexin Wang, Maozu Guo, Data fusion-based algorithm for predicting miRNA–Disease associations, 2020, 88, 14769271, 107357, 10.1016/j.compbiolchem.2020.107357 | |
21. | Ruirui Liang, Jiayang Xie, Chi Zhang, Mengying Zhang, Hai Huang, Haizhong Huo, Xin Cao, Bing Niu, Identifying Cancer Targets Based on Machine Learning Methods via Chou’s 5-steps Rule and General Pseudo Components, 2019, 19, 15680266, 2301, 10.2174/1568026619666191016155543 | |
22. | Ting Liu, Hua Tang, A Brief Survey of Machine Learning Methods in Identification of Mitochondria Proteins in Malaria Parasite, 2020, 26, 13816128, 3049, 10.2174/1381612826666200310122324 | |
23. | Zijie Sun, Shenghui Huang, Lei Zheng, Pengfei Liang, Wuritu Yang, Yongchun Zuo, ICTC-RAAC: An improved web predictor for identifying the types of ion channel-targeted conotoxins by using reduced amino acid cluster descriptors, 2020, 89, 14769271, 107371, 10.1016/j.compbiolchem.2020.107371 | |
24. | Ying Wang, Juanjuan Kang, Ning Li, Yuwei Zhou, Zhongjie Tang, Bifang He, Jian Huang, NeuroCS: A Tool to Predict Cleavage Sites of Neuropeptide Precursors, 2020, 27, 09298665, 337, 10.2174/0929866526666191112150636 | |
25. | Shanwen Sun, Hui Ding, Donghua Wang, Shuguang Han, Identifying Antifreeze Proteins Based on Key Evolutionary Information, 2020, 8, 2296-4185, 10.3389/fbioe.2020.00244 | |
26. | Nguyen Quoc Khanh Le, Fertility-GRU: Identifying Fertility-Related Proteins by Incorporating Deep-Gated Recurrent Units and Original Position-Specific Scoring Matrix Profiles, 2019, 18, 1535-3893, 3503, 10.1021/acs.jproteome.9b00411 | |
27. | Zhourun Wu, Qing Liao, Bin Liu, idenPC-MIIP: identify protein complexes from weighted PPI networks using mutual important interacting partner relation, 2021, 22, 1477-4054, 1972, 10.1093/bib/bbaa016 | |
28. | Zhe Liu, Yingli Gong, Yihang Bao, Yuanzhao Guo, Han Wang, Guan Ning Lin, TMPSS: A Deep Learning-Based Predictor for Secondary Structure and Topology Structure Prediction of Alpha-Helical Transmembrane Proteins, 2021, 8, 2296-4185, 10.3389/fbioe.2020.629937 | |
29. | Hao Lv, Fu-Ying Dao, Zheng-Xing Guan, Dan Zhang, Jiu-Xin Tan, Yong Zhang, Wei Chen, Hao Lin, iDNA6mA-Rice: A Computational Tool for Detecting N6-Methyladenine Sites in Rice, 2019, 10, 1664-8021, 10.3389/fgene.2019.00793 | |
30. | Ruiyan Hou, Jin Wu, Lei Xu, Quan Zou, Yi-Jun Wu, Computational Prediction of Protein Arginine Methylation Based on Composition–Transition–Distribution Features, 2020, 5, 2470-1343, 27470, 10.1021/acsomega.0c03972 | |
31. | Xiao-Yang Jing, Feng-Min Li, Predicting Cell Wall Lytic Enzymes Using Combined Features, 2021, 8, 2296-4185, 10.3389/fbioe.2020.627335 | |
32. | Hong-Fei Li, Xian-Fang Wang, Hua Tang, Predicting Bacteriophage Enzymes and Hydrolases by Using Combined Features, 2020, 8, 2296-4185, 10.3389/fbioe.2020.00183 | |
33. | Wei Chen, Pengmian Feng, Xiaoming Song, Hao Lv, Hao Lin, iRNA-m7G: Identifying N7-methylguanosine Sites by Fusing Multiple Features, 2019, 18, 21622531, 269, 10.1016/j.omtn.2019.08.022 | |
34. | Chao Wang, Pingping Wang, Shuguang Han, Lida Wang, Yuming Zhao, Liran Juan, FunEffector-Pred: Identification of Fungi Effector by Activate Learning and Genetic Algorithm Sampling of Imbalanced Data, 2020, 8, 2169-3536, 57674, 10.1109/ACCESS.2020.2982410 | |
35. | Yanjuan Li, Zitong Zhang, Zhixia Teng, Xiaoyan Liu, Hui Ding, PredAmyl-MLP: Prediction of Amyloid Proteins Using Multilayer Perceptron, 2020, 2020, 1748-6718, 1, 10.1155/2020/8845133 | |
36. | Wanben Zhong, Bineng Zhong, Hongbo Zhang, Ziyi Chen, Yan Chen, Identification of Anti-cancer Peptides Based on Multi-classifier System, 2020, 22, 13862073, 694, 10.2174/1386207322666191203141102 | |
37. | Xiaoqing Ru, Peigang Cao, Lihong Li, Quan Zou, Selecting Essential MicroRNAs Using a Novel Voting Method, 2019, 18, 21622531, 16, 10.1016/j.omtn.2019.07.019 | |
38. | Feifei Cui, Zilong Zhang, Quan Zou, Sequence representation approaches for sequence-based protein prediction tasks that use deep learning, 2021, 20, 2041-2649, 61, 10.1093/bfgp/elaa030 | |
39. | Muhammad Arif, Farman Ali, Saeed Ahmad, Muhammad Kabir, Zakir Ali, Maqsood Hayat, Pred-BVP-Unb: Fast prediction of bacteriophage Virion proteins using un-biased multi-perspective properties with recursive feature elimination, 2020, 112, 08887543, 1565, 10.1016/j.ygeno.2019.09.006 | |
40. | Yongxian Fan, Meijun Chen, Qingqi Zhu, lncLocPred: Predicting LncRNA Subcellular Localization Using Multiple Sequence Feature Information, 2020, 8, 2169-3536, 124702, 10.1109/ACCESS.2020.3007317 | |
41. | Changgeng Tan, Tong Wang, Wenyi Yang, Lei Deng, PredPSD: A Gradient Tree Boosting Approach for Single-Stranded and Double-Stranded DNA Binding Protein Prediction, 2019, 25, 1420-3049, 98, 10.3390/molecules25010098 | |
42. | Chunyu Wang, Jialin Li, Ying Zhang, Maozu Guo, Identification of Type VI Effector Proteins Using a Novel Ensemble Classifier, 2020, 8, 2169-3536, 75085, 10.1109/ACCESS.2020.2985111 | |
43. | Zhibin Lv, Hui Ding, Lei Wang, Quan Zou, A Convolutional Neural Network Using Dinucleotide One-hot Encoder for identifying DNA N6-Methyladenine Sites in the Rice Genome, 2021, 422, 09252312, 214, 10.1016/j.neucom.2020.09.056 | |
44. | Xiao-Yang Jing, Feng-Min Li, Identifying Heat Shock Protein Families from Imbalanced Data by Using Combined Features, 2020, 2020, 1748-670X, 1, 10.1155/2020/8894478 | |
45. | Hui Yang, Wuritu Yang, Fu-Ying Dao, Hao Lv, Hui Ding, Wei Chen, Hao Lin, A comparison and assessment of computational method for identifying recombination hotspots in Saccharomyces cerevisiae , 2020, 21, 1467-5463, 1568, 10.1093/bib/bbz123 | |
46. | Balachandran Manavalan, Md. Mehedi Hasan, Shaherin Basith, Vijayakumar Gosu, Tae-Hwan Shin, Gwang Lee, Empirical Comparison and Analysis of Web-Based DNA N4-Methylcytosine Site Prediction Tools, 2020, 22, 21622531, 406, 10.1016/j.omtn.2020.09.010 | |
47. | Zheng-Xing Guan, Shi-Hao Li, Zi-Mei Zhang, Dan Zhang, Hui Yang, Hui Ding, A Brief Survey for MicroRNA Precursor Identification Using Machine Learning Methods, 2020, 21, 13892029, 11, 10.2174/1389202921666200214125102 | |
48. | Balachandran Manavalan, Shaherin Basith, Tae Hwan Shin, Gwang Lee, Computational prediction of species-specific yeast DNA replication origin via iterative feature representation, 2020, 1467-5463, 10.1093/bib/bbaa304 | |
49. | Xian-Fang Wang, Peng Gao, Yi-Feng Liu, Hong-Fei Li, Fan Lu, Predicting Thermophilic Proteins by Machine Learning, 2020, 15, 15748936, 493, 10.2174/1574893615666200207094357 | |
50. | Chang Lu, Yingli Gong, Zhe Liu, Yuanzhao Guo, Zhiqiang Ma, Han Wang, TM-ZC: A Deep Learning-Based Predictor for the Z-Coordinate of Residues in α-Helical Transmembrane Proteins, 2020, 8, 2169-3536, 40129, 10.1109/ACCESS.2020.2976797 | |
51. | Tianyi Zhao, Donghua Wang, Yang Hu, Ningyi Zhang, Tianyi Zang, Yadong Wang, Identifying Alzheimer’s Disease-related miRNA Based on Semi-clustering, 2019, 19, 15665232, 216, 10.2174/1566523219666190924113737 | |
52. | Dan Zhang, Zhao-Chun Xu, Wei Su, Yu-He Yang, Hao Lv, Hui Yang, Hao Lin, Jinbo Xu, iCarPS: a computational tool for identifying protein carbonylation sites by novel encoded features, 2020, 1367-4803, 10.1093/bioinformatics/btaa702 | |
53. | Dan Zhang, Zheng-Xing Guan, Zi-Mei Zhang, Shi-Hao Li, Fu-Ying Dao, Hua Tang, Hao Lin, Recent Development of Computational Predicting Bioluminescent Proteins, 2020, 25, 13816128, 4264, 10.2174/1381612825666191107100758 | |
54. | Yanwen Li, Feng Pu, Yu Feng, Jinchao Ji, Hongguang Sun, Han Wang, MRMD-palm: A novel method for the identification of palmitoylated protein, 2021, 210, 01697439, 104245, 10.1016/j.chemolab.2021.104245 | |
55. | Feng-Min Li, Xiao-Wei Gao, Predicting Gram-Positive Bacterial Protein Subcellular Location by Using Combined Features, 2020, 2020, 2314-6133, 1, 10.1155/2020/9701734 | |
56. | Zhiyu Tao, Yanjuan Li, Zhixia Teng, Yuming Zhao, Hui Ding, A Method for Identifying Vesicle Transport Proteins Based on LibSVM and MRMD, 2020, 2020, 1748-6718, 1, 10.1155/2020/8926750 | |
57. | Duyen Thi Do, Nguyen Quoc Khanh Le, A sequence-based approach for identifying recombination spots in Saccharomyces cerevisiae by using hyper-parameter optimization in FastText and support vector machine, 2019, 194, 01697439, 103855, 10.1016/j.chemolab.2019.103855 | |
58. | Chunyu Wang, Jialin Li, Xiaoyan Liu, Maozu Guo, Predicting Sub-Golgi Apparatus Resident Protein With Primary Sequence Hybrid Features, 2020, 8, 2169-3536, 4442, 10.1109/ACCESS.2019.2962821 | |
59. | Qingwen Li, Benzhi Dong, Donghua Wang, Sui Wang, Identification of Secreted Proteins From Malaria Protozoa With Few Features, 2020, 8, 2169-3536, 89793, 10.1109/ACCESS.2020.2994206 | |
60. | Pengmian Feng, Weiwei Liu, Cong Huang, Zhaohui Tang, Classifying the superfamily of small heat shock proteins by using g-gap dipeptide compositions, 2021, 167, 01418130, 1575, 10.1016/j.ijbiomac.2020.11.111 | |
61. | Hongfei Li, Haoze Du, Xianfang Wang, Peng Gao, Yifeng Liu, Weizhong Lin, Remarks on Computational Method for Identifying Acid and Alkaline Enzymes, 2020, 26, 13816128, 3105, 10.2174/1381612826666200617170826 | |
62. | Shahid Akbar, Salman Khan, Farman Ali, Maqsood Hayat, Muhammad Qasim, Sarah Gul, iHBP-DeepPSSM: Identifying hormone binding proteins using PsePSSM based evolutionary features and deep learning approach, 2020, 204, 01697439, 104103, 10.1016/j.chemolab.2020.104103 | |
63. | He Zhuang, Ying Zhang, Shuo Yang, Liang Cheng, Shu-Lin Liu, A Mendelian Randomization Study on Infant Length and Type 2 Diabetes Mellitus Risk, 2019, 19, 15665232, 224, 10.2174/1566523219666190925115535 | |
64. | Xin Gao, Donghua Wang, Jun Zhang, Qing Liao, Bin Liu, iRBP-Motif-PSSM: Identification of RNA-Binding Proteins Based on Collaborative Learning, 2019, 7, 2169-3536, 168956, 10.1109/ACCESS.2019.2952621 | |
65. | Zihao Yan, Dong Chen, Zhixia Teng, Donghua Wang, Yanjuan Li, SMOPredT4SE: An Effective Prediction of Bacterial Type IV Secreted Effectors Using SVM Training With SMO, 2020, 8, 2169-3536, 25570, 10.1109/ACCESS.2020.2971091 | |
66. | Dan Zhang, Hua-Dong Chen, Hasan Zulfiqar, Shi-Shi Yuan, Qin-Lai Huang, Zhao-Yue Zhang, Ke-Jun Deng, Watshara Shoombuatong, iBLP: An XGBoost-Based Predictor for Identifying Bioluminescent Proteins, 2021, 2021, 1748-6718, 1, 10.1155/2021/6664362 | |
67. | Zi-Mei Zhang, Jiu-Xin Tan, Fang Wang, Fu-Ying Dao, Zhao-Yue Zhang, Hao Lin, Early Diagnosis of Hepatocellular Carcinoma Using Machine Learning Method, 2020, 8, 2296-4185, 10.3389/fbioe.2020.00254 | |
68. | Zifan Guo, Pingping Wang, Zhendong Liu, Yuming Zhao, Discrimination of Thermophilic Proteins and Non-thermophilic Proteins Using Feature Dimension Reduction, 2020, 8, 2296-4185, 10.3389/fbioe.2020.584807 | |
69. | Rajiv G. Govindaraj, Sathiyamoorthy Subramaniyam, Balachandran Manavalan, Extremely-randomized-tree-based Prediction of N6-methyladenosine Sites inSaccharomyces cerevisiae, 2020, 21, 13892029, 26, 10.2174/1389202921666200219125625 | |
70. | Sola Gbenro, Kyle Hippe, Renzhi Cao, 2020, HMMeta, 9781450379649, 1, 10.1145/3388440.3414702 | |
71. | Yixiao Zhai, Yu Chen, Zhixia Teng, Yuming Zhao, Identifying Antioxidant Proteins by Using Amino Acid Composition and Protein-Protein Interactions, 2020, 8, 2296-634X, 10.3389/fcell.2020.591487 | |
72. | Xianhai Li, Qiang Tang, Hua Tang, Wei Chen, Identifying Antioxidant Proteins by Combining Multiple Methods, 2020, 8, 2296-4185, 10.3389/fbioe.2020.00858 | |
73. | Bing Rao, Lichao Zhang, Guoying Zhang, ACP-GCN: The Identification of Anticancer Peptides Based on Graph Convolution Networks, 2020, 8, 2169-3536, 176005, 10.1109/ACCESS.2020.3023800 | |
74. | Ni Kou, Wenyang Zhou, Yuzhu He, Xiaoxia Ying, Songling Chai, Tao Fei, Wenqi Fu, Jiaqian Huang, Huiying Liu, A Mendelian Randomization Analysis to Expose the Causal Effect of IL-18 on Osteoporosis Based on Genome-Wide Association Study Data, 2020, 8, 2296-4185, 10.3389/fbioe.2020.00201 | |
75. | Peng Xu, Geng Zhao, Zheng Kou, Gang Fang, Wenbin Liu, Classification of Cancers Based on a Comprehensive Pathway Activity Inferred by Genes and Their Interactions, 2020, 8, 2169-3536, 30515, 10.1109/ACCESS.2020.2973220 | |
76. | Shi-Hao Li, Zheng-Xing Guan, Dan Zhang, Zi-Mei Zhang, Jian Huang, Wuritu Yang, Hao Lin, Recent Advancement in Predicting Subcellular Localization of Mycobacterial Protein with Machine Learning Methods, 2020, 16, 15734064, 605, 10.2174/1573406415666191004101913 | |
77. | Han Luo, Donghua Wang, Juan Liu, Ying Ju, Zhe Jin, A Framework Integrating Heterogeneous Databases for the Completion of Gene Networks, 2019, 7, 2169-3536, 168859, 10.1109/ACCESS.2019.2954994 | |
78. | Zhibin Lv, Feifei Cui, Quan Zou, Lichao Zhang, Lei Xu, Anticancer peptides prediction with deep representation learning features, 2021, 1467-5463, 10.1093/bib/bbab008 | |
79. | Chang Lu, Zhe Liu, Bowen Kan, Yingli Gong, Zhiqiang Ma, Han Wang, TMP-SSurface: A Deep Learning-Based Predictor for Surface Accessibility of Transmembrane Protein Residues, 2019, 9, 2073-4352, 640, 10.3390/cryst9120640 | |
80. | Yideng Cai, Jiacheng Wang, Lei Deng, SDN2GO: An Integrated Deep Learning Model for Protein Function Prediction, 2020, 8, 2296-4185, 10.3389/fbioe.2020.00391 | |
81. | Hao Wang, Qilemuge Xi, Pengfei Liang, Lei Zheng, Yan Hong, Yongchun Zuo, IHEC_RAAC: a online platform for identifying human enzyme classes via reduced amino acid cluster strategy, 2021, 53, 0939-4451, 239, 10.1007/s00726-021-02941-9 | |
82. | Shanwen Sun, Chunyu Wang, Hui Ding, Quan Zou, Machine learning and its applications in plant molecular studies, 2020, 19, 2041-2657, 40, 10.1093/bfgp/elz036 | |
83. | Yichen Guo, Ke Yan, Hao Wu, Bin Liu, ReFold-MAP: Protein remote homology detection and fold recognition based on features extracted from profiles, 2020, 611, 00032697, 114013, 10.1016/j.ab.2020.114013 | |
84. | Guilin Li, Xing Gao, The Feature Compression Algorithms for Identifying Cytokines Based on CNT Features, 2020, 8, 2169-3536, 83645, 10.1109/ACCESS.2020.2989749 | |
85. | Qilemuge Xi, Hao Wang, Liuxi Yi, Jian Zhou, Yuchao Liang, Xiaoqing Zhao, Yongchun Zuo, Lei Chen, ANPrAod: Identify Antioxidant Proteins by Fusing Amino Acid Clustering Strategy and N -Peptide Combination, 2021, 2021, 1748-6718, 1, 10.1155/2021/5518209 | |
86. | Lesong Wei, Xiucai Ye, Yuyang Xue, Tetsuya Sakurai, Leyi Wei, ATSE: a peptide toxicity predictor by exploiting structural and evolutionary information based on graph neural network and attention mechanism, 2021, 1467-5463, 10.1093/bib/bbab041 | |
87. | Hasan Zulfiqar, Rida Sarwar Khan, Farwa Hassan, Kyle Hippe, Cassandra Hunt, Hui Ding, Xiao-Ming Song, Renzhi Cao, Computational identification of N4-methylcytosine sites in the mouse genome with machine-learning method, 2021, 18, 1551-0018, 3348, 10.3934/mbe.2021167 | |
88. | Yanwen Li, Feng Pu, Jingru Wang, Zhiguo Zhou, Chunhua Zhang, Fei He, Zhiqiang Ma, Jingbo Zhang, Machine Learning Methods in Prediction of Protein Palmitoylation Sites: A Brief Review, 2021, 27, 13816128, 2189, 10.2174/1381612826666201112142826 | |
89. | Shulin Zhao, Ying Ju, Xiucai Ye, Jun Zhang, Shuguang Han, Bioluminescent Proteins Prediction with Voting Strategy, 2021, 16, 15748936, 240, 10.2174/1574893615999200601122328 | |
90. | Yinuo Lyu, Zhen Zhang, Jiawei Li, Wenying He, Yijie Ding, Fei Guo, iEnhancer-KL: A Novel Two-Layer Predictor for Identifying Enhancers by Position Specific of Nucleotide Composition, 2021, 18, 1545-5963, 2809, 10.1109/TCBB.2021.3053608 | |
91. | Hongdi Pei, Jiayu Li, Shuhan Ma, Jici Jiang, Mingxin Li, Quan Zou, Zhibin Lv, Identification of Thermophilic Proteins Based on Sequence-Based Bidirectional Representations from Transformer-Embedding Features, 2023, 13, 2076-3417, 2858, 10.3390/app13052858 | |
92. | Hasan Zulfiqar, Qin-Lai Huang, Hao Lv, Zi-Jie Sun, Fu-Ying Dao, Hao Lin, Deep-4mCGP: A Deep Learning Approach to Predict 4mC Sites in Geobacter pickeringii by Using Correlation-Based Feature Selection Technique, 2022, 23, 1422-0067, 1251, 10.3390/ijms23031251 | |
93. | Ting Liu, Jiamao Chen, Qian Zhang, Kyle Hippe, Cassandra Hunt, Thu Le, Renzhi Cao, Hua Tang, The Development of Machine Learning Methods in Discriminating Secretory Proteins of Malaria Parasite, 2022, 29, 09298673, 807, 10.2174/0929867328666211005140625 | |
94. | Jing Guo, TCN-HBP: A Deep Learning Method for Identifying Hormone-Binding Proteins from Amino Acid Sequences Based on a Temporal Convolution Neural Network, 2021, 2025, 1742-6588, 012002, 10.1088/1742-6596/2025/1/012002 | |
95. | Chaolu Meng, Jin Wu, Fei Guo, Benzhi Dong, Lei Xu, CWLy-pred: A novel cell wall lytic enzyme identifier based on an improved MRMD feature selection method, 2020, 112, 08887543, 4715, 10.1016/j.ygeno.2020.08.015 | |
96. | Zhixia Teng, Zitong Zhang, Zhen Tian, Yanjuan Li, Guohua Wang, ReRF-Pred: predicting amyloidogenic regions of proteins based on their pseudo amino acid composition and tripeptide composition, 2021, 22, 1471-2105, 10.1186/s12859-021-04446-4 | |
97. | Peiran Jiang, Wanshan Ning, Yunshu Shi, Chuan Liu, Saijun Mo, Haoran Zhou, Kangdong Liu, Yaping Guo, FSL-Kla: A few-shot learning-based multi-feature hybrid system for lactylation site prediction, 2021, 19, 20010370, 4497, 10.1016/j.csbj.2021.08.013 | |
98. | Mujiexin Liu, Hui Chen, Dong Gao, Cai-Yi Ma, Zhao-Yue Zhang, Balachandran Manavalan, Identification of Helicobacter pylori Membrane Proteins Using Sequence-Based Features, 2022, 2022, 1748-6718, 1, 10.1155/2022/7493834 | |
99. | Yi-Wei Zhao, Shihua Zhang, Hui Ding, Recent Development of Machine Learning Methods in Sumoylation Sites Prediction, 2022, 29, 09298673, 894, 10.2174/0929867328666210915112030 | |
100. | Juexin Wang, Yan Wang, Towards Machine Learning in Molecular Biology, 2020, 17, 1551-0018, 2822, 10.3934/mbe.2020156 | |
101. | Yuning Yang, Jiawen Yu, Zhe Liu, Xi Wang, Han Wang, Zhiqiang Ma, Dong Xu, An Improved Topology Prediction of Alpha-Helical Transmembrane Protein Based on Deep Multi-Scale Convolutional Neural Network, 2022, 19, 1545-5963, 295, 10.1109/TCBB.2020.3005813 | |
102. | Shihu Jiao, Zheng Chen, Lichao Zhang, Xun Zhou, Lei Shi, ATGPred-FL: sequence-based prediction of autophagy proteins with feature representation learning, 2022, 54, 0939-4451, 799, 10.1007/s00726-022-03145-5 | |
103. | Changli Feng, Haiyan Wei, Deyun Yang, Bin Feng, Zhaogui Ma, Shuguang Han, Quan Zou, Hua Shi, ORS‐Pred: An optimized reduced scheme‐based identifier for antioxidant proteins, 2021, 21, 1615-9853, 2100017, 10.1002/pmic.202100017 | |
104. | Hongliang Zou, Zhijian Yin, m7G-DPP: Identifying N7-methylguanosine sites based on dinucleotide physicochemical properties of RNA, 2021, 279, 03014622, 106697, 10.1016/j.bpc.2021.106697 | |
105. | Yuxin Gong, Bo Liao, Dejun Peng, Quan Zou, Accurate Prediction and Key Feature Recognition of Immunoglobulin, 2021, 11, 2076-3417, 6894, 10.3390/app11156894 | |
106. | Yuxin Guo, Liping Hou, Wen Zhu, Peng Wang, Prediction of Hormone-Binding Proteins Based on K-mer Feature Representation and Naive Bayes, 2021, 12, 1664-8021, 10.3389/fgene.2021.797641 | |
107. | Lesong Wei, Xiucai Ye, Tetsuya Sakurai, Zengchao Mu, Leyi Wei, Pier Luigi Martelli, ToxIBTL: prediction of peptide toxicity based on information bottleneck and transfer learning, 2022, 38, 1367-4803, 1514, 10.1093/bioinformatics/btac006 | |
108. | Yihang Bao, Weixi Wang, Minglong Dong, Fei He, Han Wang, 2021, Discover the Binding Domain of Transmembrane Proteins Based on Structural Universality, 978-1-6654-0126-5, 5, 10.1109/BIBM52615.2021.9669493 | |
109. | Wen Zhu, Yuxin Guo, Quan Zou, Prediction of presynaptic and postsynaptic neurotoxins based on feature extraction, 2021, 18, 1551-0018, 5943, 10.3934/mbe.2021297 | |
110. | Hasan Zulfiqar, Shi-Shi Yuan, Qin-Lai Huang, Zi-Jie Sun, Fu-Ying Dao, Xiao-Long Yu, Hao Lin, Identification of cyclin protein using gradient boost decision tree algorithm, 2021, 19, 20010370, 4123, 10.1016/j.csbj.2021.07.013 | |
111. | Jun Zhang, Qingcai Chen, Bin Liu, DeepDRBP-2L: A New Genome Annotation Predictor for Identifying DNA-Binding Proteins and RNA-Binding Proteins Using Convolutional Neural Network and Long Short-Term Memory, 2021, 18, 1545-5963, 1451, 10.1109/TCBB.2019.2952338 | |
112. | Pengmian Feng, Lijing Feng, Recent Advances on Antioxidant Identification Based on Machine Learning Methods, 2020, 21, 13892002, 804, 10.2174/1389200221666200719001449 | |
113. | Xiaoping Min, Fengqing Lu, Chunyan Li, Sequence-Based Deep Learning Frameworks on Enhancer-Promoter Interactions Prediction, 2021, 27, 13816128, 1847, 10.2174/1381612826666201124112710 | |
114. | Yu-Hao Wang, Yu-Fei Zhang, Ying Zhang, Zhi-Feng Gu, Zhao-Yue Zhang, Hao Lin, Ke-Jun Deng, Identification of adaptor proteins using the ANOVA feature selection technique, 2022, 208, 10462023, 42, 10.1016/j.ymeth.2022.10.008 | |
115. | Shihu Jiao, Quan Zou, Huannan Guo, Lei Shi, iTTCA-RF: a random forest predictor for tumor T cell antigens, 2021, 19, 1479-5876, 10.1186/s12967-021-03084-x | |
116. | Bosheng Song, Zimeng Li, Xuan Lin, Jianmin Wang, Tian Wang, Xiangzheng Fu, Pretraining model for biological sequence data, 2021, 20, 2041-2649, 181, 10.1093/bfgp/elab025 | |
117. | Peijie Zheng, Yue Qi, Xueyong Li, Yuewu Liu, Yuhua Yao, Guohua Huang, A capsule network-based method for identifying transcription factors, 2022, 13, 1664-302X, 10.3389/fmicb.2022.1048478 | |
118. | Hongliang Zou, iAHTP-LH: Integrating Low-Order and High-Order Correlation Information for Identifying Antihypertensive Peptides, 2022, 28, 1573-3904, 10.1007/s10989-022-10414-0 | |
119. | Hui Zhang, Qin Chen, Bing Niu, Risk Assessment of Veterinary Drug Residues in Meat Products, 2020, 21, 13892002, 779, 10.2174/1389200221999200820164650 | |
120. | Ke Yan, Jie Wen, Jin-Xing Liu, Yong Xu, Bin Liu, Protein Fold Recognition by Combining Support Vector Machines and Pairwise Sequence Similarity Scores, 2021, 18, 1545-5963, 2008, 10.1109/TCBB.2020.2966450 | |
121. | Hongliang Zou, Fan Yang, Zhijian Yin, Identification of tumor homing peptides by utilizing hybrid feature representation, 2022, 0739-1102, 1, 10.1080/07391102.2022.2049368 | |
122. | Hasan Zulfiqar, Zhiling Guo, Bakanina Kissanga Grace-Mercure, Zhao-Yue Zhang, Hui Gao, Hao Lin, Yun Wu, Empirical Comparison and Recent Advances of Computational Prediction of Hormone Binding Proteins Using Machine Learning Methods, 2023, 20010370, 10.1016/j.csbj.2023.03.024 | |
123. | Hongliang Zou, iHBPs-VWDC: variable-length window-based dynamic connectivity approach for identifying hormone-binding proteins, 2023, 0739-1102, 1, 10.1080/07391102.2023.2283150 | |
124. | Ali Ghulam, Zar Nawab Khan Swati, Farman Ali, Saima Tunio, Nida Jabeen, Natasha Iqbal, DeepImmuno-PSSM: Identification of Immunoglobulin based on Deep learning and PSSM-Profiles, 2023, 11, 2308-8168, 54, 10.21015/vtcs.v11i1.1396 | |
125. | A. Sherly Alphonse, N. Ani Brown Mary, Classification of anti-oxidant proteins using novel physiochemical and conjoint-quad (PCQ) feature composition, 2023, 83, 1573-7721, 48831, 10.1007/s11042-023-17498-w | |
126. | Zhi-Feng Gu, Yu-Duo Hao, Tian-Yu Wang, Pei-Ling Cai, Yang Zhang, Ke-Jun Deng, Hao Lin, Hao Lv, Prediction of blood–brain barrier penetrating peptides based on data augmentation with Augur, 2024, 22, 1741-7007, 10.1186/s12915-024-01883-4 | |
127. | Zhibin Lv, Mingxuan Wei, Hongdi Pei, Shiyu Peng, Mingxin Li, Liangzhen Jiang, PTSP-BERT: Predict the thermal stability of proteins using sequence-based bidirectional representations from transformer-embedded features, 2025, 185, 00104825, 109598, 10.1016/j.compbiomed.2024.109598 | |
128. | Ali Ghulam, Muhammad Arif, Ahsanullah Unar, Maha A. Thafar, Somayah Albaradei, Apilak Worachartcheewan, StackAHTPs: An explainable antihypertensive peptides identifier based on heterogeneous features and stacked learning approach, 2025, 19, 1751-8849, 10.1049/syb2.70002 | |
129. | Rui Li, Junwen Yu, Dongxin Ye, Shanghua Liu, Hongqi Zhang, Hao Lin, Juan Feng, Kejun Deng, Conotoxins: Classification, Prediction, and Future Directions in Bioinformatics, 2025, 17, 2072-6651, 78, 10.3390/toxins17020078 | |
130. | Firuz Kamalov, Hana Sulieman, Ayman Alzaatreh, Maher Emarly, Hasna Chamlal, Murodbek Safaraliev, Mathematical Methods in Feature Selection: A Review, 2025, 13, 2227-7390, 996, 10.3390/math13060996 | |
131. | Dheeraj Raya, Vincent Peta, Alain Bomgni, Shiva Aryal, Tuyen Duc Do, Kalimuthu Jawaharraj, David R. Salem, Venkataramana Gadhamshetty, Saurabh Sudha Dhiman, Etienne Z. Gnimpieba, Prediction and validation of nanowire proteins in Oleidesulfovibrio alaskensis G20 using machine learning and feature engineering, 2025, 20010370, 10.1016/j.csbj.2025.04.022 |
Parameter | Description | Unit | Value |
β | Inverse of the cytoplasmic density | Lg−1 | 0.003 |
kM | Rate constant of the metabolic macroreaction | h−1 | 4.32 |
KM | Half-saturation constant of the metabolic macroreaction | gL−1 | 33.33 |
kR | Rate constant of the protein synthesis reaction | h−1 | 3.6 |
KR | Half-saturation constant of the protein synthesis reaction | gL−1 | 1 |
kX | Rate constant of the metabolite synthesis reaction | h−1 | 0.5 |
KX | Half-saturation constant of the metabolite synthesis reaction | gL−1 | 1 |
Parameter | Description | Unit | Value |
β | Inverse of the cytoplasmic density | Lg−1 | 0.003 |
kM | Rate constant of the metabolic macroreaction | h−1 | 4.32 |
KM | Half-saturation constant of the metabolic macroreaction | gL−1 | 33.33 |
kR | Rate constant of the protein synthesis reaction | h−1 | 3.6 |
KR | Half-saturation constant of the protein synthesis reaction | gL−1 | 1 |
kX | Rate constant of the metabolite synthesis reaction | h−1 | 0.5 |
KX | Half-saturation constant of the metabolite synthesis reaction | gL−1 | 1 |