
Citation: Thomas Rauen, Rose Tanui, Christof Grewer. Structural and functional dynamics of Excitatory Amino Acid Transporters (EAAT)[J]. AIMS Molecular Science, 2014, 1(3): 99-125. doi: 10.3934/molsci.2014.3.99
[1] | Yuzhuo Shi, Huijie Zhang, Zhisheng Li, Kun Hao, Yonglei Liu, Lu Zhao . Path planning for mobile robots in complex environments based on improved ant colony algorithm. Mathematical Biosciences and Engineering, 2023, 20(9): 15568-15602. doi: 10.3934/mbe.2023695 |
[2] | Teng Fei, Xinxin Wu, Liyi Zhang, Yong Zhang, Lei Chen . Research on improved ant colony optimization for traveling salesman problem. Mathematical Biosciences and Engineering, 2022, 19(8): 8152-8186. doi: 10.3934/mbe.2022381 |
[3] | Shihong Yin, Qifang Luo, Yanlian Du, Yongquan Zhou . DTSMA: Dominant Swarm with Adaptive T-distribution Mutation-based Slime Mould Algorithm. Mathematical Biosciences and Engineering, 2022, 19(3): 2240-2285. doi: 10.3934/mbe.2022105 |
[4] | Yijie Zhang, Yuhang Cai . Adaptive dynamic self-learning grey wolf optimization algorithm for solving global optimization problems and engineering problems. Mathematical Biosciences and Engineering, 2024, 21(3): 3910-3943. doi: 10.3934/mbe.2024174 |
[5] | Zhen Zhou, Chenchen Geng, Buhu Qi, Aiwen Meng, Jinzhuang Xiao . Research and experiment on global path planning for indoor AGV via improved ACO and fuzzy DWA. Mathematical Biosciences and Engineering, 2023, 20(11): 19152-19173. doi: 10.3934/mbe.2023846 |
[6] | Rong Zheng, Heming Jia, Laith Abualigah, Shuang Wang, Di Wu . An improved remora optimization algorithm with autonomous foraging mechanism for global optimization problems. Mathematical Biosciences and Engineering, 2022, 19(4): 3994-4037. doi: 10.3934/mbe.2022184 |
[7] | Rongmei Geng, Renxin Ji, Shuanjin Zi . Research on task allocation of UAV cluster based on particle swarm quantization algorithm. Mathematical Biosciences and Engineering, 2023, 20(1): 18-33. doi: 10.3934/mbe.2023002 |
[8] | Yuting Liu, Hongwei Ding, Zongshan Wang, Gushen Jin, Bo Li, Zhijun Yang, Gaurav Dhiman . A chaos-based adaptive equilibrium optimizer algorithm for solving global optimization problems. Mathematical Biosciences and Engineering, 2023, 20(9): 17242-17271. doi: 10.3934/mbe.2023768 |
[9] | Xiaoding Meng, Hecheng Li, Anshan Chen . Multi-strategy self-learning particle swarm optimization algorithm based on reinforcement learning. Mathematical Biosciences and Engineering, 2023, 20(5): 8498-8530. doi: 10.3934/mbe.2023373 |
[10] | Xue Li, Lei Wang . Application of improved ant colony optimization in mobile robot trajectory planning. Mathematical Biosciences and Engineering, 2020, 17(6): 6756-6774. doi: 10.3934/mbe.2020352 |
In the last decades, combinatorial optimization problems have attract the interest of researchers due to the increase of dynamic environments applied on these problems, and researchers have developed mechanisms and methods to help optimization algorithms to adapt to changes that exist during the execution of these algorithms.
These problems can be represented by many production and services problems, including the reduction of the product's cost, improvement of logistic and company's profit. Furthermore, transport cost minimization is one of the cost reduction methods in which the mercancy is transported from an origin place to the one or several destinations with minimum cost [1].
Even defining Dynamic Combinatorial Optimization Problems (DCOP's or Dynamic COP's) is a challenging task due to applications in real problems such as transportation of products where there are a lot of not considered variables which cannot be identified until implementation on real cases. In general, researchers usually define optimization problems that change over time as Dynamic Problems, Time-Dependent Problems, or Dynamic Optimization Problems [2]. In existing metaheuristics studies, dynamic problems are defined as a sequence of static problems linked up by some dynamic rules or as a problem that have time dependent parameters in its mathematical model.
Usually, for static Combinatorial Optimization Problems (COP's), the goal of metaheuristics is to find the global optimum as fast and precise as possible, considering performance measures such as the convergence speed and the rate to get the optimum over multiple runs.
However, the goal of algorithms to solve DCOP's turns from to find the global optimum as fast as possible [3], to track the optimum as close as possible to real time on dynamic changes; in several cases, the algorithm needs to detect these changes, and then, it needs to track the local or global optimum. In addition, on environments where there exists a correlation between changes, the optimization algorithm needs to learn from its previous experience as a feedback to improve the search in the new solution space. Otherwise, the optimization process after each change will be explained like the process to solve different problems starting from the old population.
Heuristics and metaheuristics are methods that have been used to solve several problems in logistics as cost reduction in supply chain distribution [4,5] and facility location problems [6], even in manufacturing like optimization of manufacturing systems [7,8] and manufacturing models [9].
This research focuses on theoretical dynamic combinatorial problems to demonstrate the ability of a proposed mechanism to adapt in dynamic environments which could describe the behavior of a real problem. In this case, the mechanism was implemented in genetic algorithms to solve a travelling salesman problem which can be generalized in several problems, such as vehicle routing problem which is used in several areas of logistics and manufacturing. These mechanism is tested in other theoretical problems like One Max and 3-SAT problem to demonstrate the capability to solve binary problems, which are considered as simple problems but genetic algorithms search have a little complication to solve them due to the behavior and evolutive process of mutation and crossover operators.
Many dynamic optimization problems have been used in literature [10], many of them have features and can be classified into different groups based on some criteria:
● Time-Linkage: Whether the future behavior of the problem depends on the current solution found by an algorithm or not.
● Predictability: Whether the generated changes are predictable or not.
● Visibility: Whether the changes are visible to the optimization algorithm and whether the changes can be detected by using just a few detectors.
● Constrained problem: Whether the problem is constrained or not.
● Number of objectives: Whether the problem has single objective or multiple objectives.
● Types of changes: Detailed explanation of how changes occur in the search space.
● Cyclicity: Whether the changes are cyclic/recurrent in the search space or not.
● Periodicity: Whether the changes area periodical or not.
● Factors that change: Changes may involve parameters of objective functions, variables domain, variables number, constraints, among others.
The problems defined in this research are Travelling Salesman Problem (TSP) and One-Max problem, which have been considered in their static and dynamic form to extend the analysis of several assessments between a set of configurations for genetic algorithm and a self-adaptive mechanism proposed implemented in genetic algorithm to improve the search of solutions in dynamic environments. The algorithm will be explained in next section.
This problem consists of a logical propositional formula in n variables and the requirement to find a value (true or false) for each variable that makes the formula true. This problem has N = 2n assignments. For k-SAT, the formula consists of a conjunction of clauses and each clause is a disjunction of k variables, any of which may be negated. For k≥3 these problems are NP-complete. An example of such a clause for k = 3, with the third variable negated, is V1 OR V2 OR (NOT V3), which is false for exactly one assignment for these variables: V1 = False, V2 = False, V3 = True.
3-SAT is one of Karp's 21 NP-complete problems [11], using Cook-Levin theorem to show that there is a polynomial time many-one reduction from the boolean satisfiability problem to each of 21 combinatorial and graph theoretical computational problems, thereby showing that they are all NP-complete. The instances that were used to test this problem were obtained from DIMACS-SATLIB Benchmark Problems.
To generate the dynamism in the 3-SAT problem, a strategy of insertion and elimination of conjunctions is used. At the beginning of the execution of the algorithm, only one strategy is selected to use it for the whole execution, the strategy can be generate a random conjunction with random variables or delete a random conjunction in the set of existing conjunctions. Since the original instance is the one of 3SATLib, when new conjunctions are inserted, it is not known in what will be the reaction of the new solutions. In the case of the elimination is an easier in this way because it does not increase the complexity of the problem, on the contrary, it reduces it.
The One-Max Problem (or Bit Counting) is a simple problem consisting in maximizing the number of ones of a bit-string. Formally, this problem can be described as finding a string $\vec x = {x_1}, {x_2}, \ldots, {x_n} $ with ${x_i} \in 1 $, that maximizes the following equation:
$ F(\vec{x}) = \sum x_{i}^{n} i = 1 $ | (1) |
Many authors, studying this problem, define this problem as a simple unimodal function and for a simple hill-climber using a traditional bit-flipping neighborhood search operator. However, a GA does not search over this simple 'Hamming' landscape. It is easy to analyze the landscape induced by a traditional NS operator, but the position is a little more complicated for the operators normally used by a GA, this is the reason of why this problem is important.
This is the simplest problem that can be selected, to convert this problem of a static environment to a dynamic environment is needed to change the search, it means, if the objective function is to find most of 1's in the solution, the change will be to find the most of 0s in the solution. This causes that the algorithms start to converge in determined generations, and when the change appears, all the solutions that have a good fitness, they are changing by the worst solution because they have the worst fitness in that moment.
TSP can be formulated as an integer linear program. Label the cities with the numbers $1, \ldots, n $ and define ${X_{i, j}} $, 1 if the path goes from city $i $ to city $ j$, and 0 otherwise. For $ {\rm{i}} = 1, \ldots, n$ let ${u_i} $ a dummy variable, and ${c_{i, j}} $ the distance from city $i $ to city $ j$. Then TSP can be written as:
$ {min \sum\limits_{i = 1}^n {\sum\limits_{j \ne i, j = 1}^n {{c_{ij}}} } {x_{ij}}} $ | (2) |
$ {{x_{i, j}} \in \{ 0, 1\} \quad i, j = 1, \ldots , n} $ | (3) |
$ {{u_i} \in {\rm{Z}}\quad i = 1, \ldots , n} $ | (4) |
$ {\sum\limits_{i = 1, i \ne j}^n {{x_{ij}}} = 1}~~~{i = 1, \ldots , n} $ | (5) |
$ {\sum\limits_{j = 1, j \ne i}^n {{x_{ij}}} = 1}~~~{i = 1, \ldots , n} $ | (6) |
$ {{u_i} - {u_j} + n{x_{ij}} \le n - 1}~~~~{2 \le i \ne j \le n} $ | (7) |
$ {0 \le {u_i} \le n - 1}~~{2 \le i \le n} $ | (8) |
DTSP have several real applications, especially in the optimization of dynamic networks, like planning and designing networks, load-balance routing, traffic management, among others.
An example of a DTSP is defined in [12] as a TSP with a dynamic cost matrix as:
$ D(t) = d_{i, j}(t)_{n * n} $ | (9) |
where: $ {d_{i, j}}\left(t \right)$ is the distance cost from city $i $ to city $ j$, and $ n$ is the number of cities. DTSP can be defined as $f\left({x, t} \right) $, and the objective of DTSP is to find a minimum-cost route with all cities at time $t $. It can be described as follow:
$ f(x, t) = {Min}\left(\sum\limits_{i = 1}^{n} d_{t i, t i+1}(t)\right. $ | (10) |
where: $ x_{i} \in 1, 2, \ldots, n$ denotes the $ i$ -th city $i $ in the solution such that $x_{n}+1 = x 1 $ and, if $ i$ is different of $ j$, and $ {x_i}$ different of ${x_j} $.
This problem have been used in manufacturing problems such as distribution and routing problems [13,14,15] and in some cases to optimize facilities [16] and machines location [17].
Genetic Algorithms (GA's) have been developed in the 1970s by J. Holland to understand the adaptive processes of natural systems [18]. Then, they have been applied to optimization and machine learning in the 1980s [19,20].
Genetic algorithms are defined in four steps: the first one is focused in the creation of a random population of candidate solutions according with the objective function. The second step is focused on each individual is evaluated in objective function. The third one, parents will be selected of the population, then is applied the crossover operator to these parents to build new solutions called children, then, the mutation operator is applied to these children. In the last step, children population will be the population for the new generation of parents, and this process will be repeated until a stop criterion defined [19]. This process is in Figure 1.
It is well known that the values of parameter settings for metaheuristics has a great impact on their performance and this has attracted considerable interest in various mechanisms that in some way attempt to automatically adjust the algorithms parameters for a given problem. Adaptive and self-adaptive methods are developed to solve problems that have dynamic features.
In [20], there is a classification according to levels of adaptation or the mechanics on these methods. In mechanics classification exist: static, deterministic, adaptive and self-adaptive. In static methods the users need to define the parameters values manually; in deterministic, a heuristic function is defined to select the values; in adaptive, a heuristic function takes advantage of some feedback while the algorithm is running; and in self-adaptive, the parameters are encoded as part of individuals allowing the algorithm to operate directly on the parameters.
In self-adapt levels exists four classes: environment, population, individual and component. On environment level, the changes will be on fitness function; on population level, the changes will be on individuals set; on individual level, changes are directly related with each individual; and in component level, changes are made on each component of each individual, for example genes in genetic algorithm.
Adaptation and self-adaptation methods have been explored on genetic algorithm components, but these mechanisms do not have relevant improvements on all of them. The most considerable improvement is in mutation and crossover operations, where new genetic operators are generated to preserve the population distribution and a degree of diversity in future off-springs. Ingo Rechenberg proposed the first deterministic adaptation rule for mutation parameter, called the "1/5 rule", that indicates in each generation must exist 1/5 of total genes mutated. A proposal of a crossover operator is named Simulated Binary Reproduction (SBX), it is presented by Deb and Beyer in [22] where SBX creates children solutions in proportion to the difference in parent solution, their purpose is to obtain results as evolutionary strategies but as genetic algorithms. Another research [22] uses a genetic operator known as Multiple Crosses Per Couple (MCPC) in which the quantity of crossovers allowed per individual is encoded in the chromosome. Similar results are obtained via the design of self-adaptive crossover operator, specifying rules such as preservation of the statistical moments in distribution of the population and the diversity degree in future on-springs in [23], in this case, Uniformly Unimodal Distribution Crossover (UNDX) is defined.
Others researches [24] have developed the idea of Immigrants Schemes in genetic algorithms to solve dynamic routing problems, these schemes are based in random immigrants approach, which is a natural and simple way to maintain the diversity level of population through replacing some individuals of the current population with random individuals, called random immigrants, every generation. Usually there are two strategies to select individuals in the population that should be replaced: replacing random or the worst individuals. Authors [24] have used Memory Schemes at the same time, in this case, memory store useful information from the current environment, either implicitly through redundant representations or explicitly by storing best solutions of current population in an extra memory. The stored information can be reused later in new environments. Finally there are algorithms using hybrid immigrants and memory schemes.
The proposed mechanism is based on the idea of a genotype-phenotype mapping function which generates random values for each parameter of a Weibull probability distribution and finally through this distribution is obtained the value of each parameter of genetic algorithm. This mechanism belongs to self-adaptive methods in the individual level described in previous section. Before to explain how the mechanism works, genotype-phenotype mapping idea needs to be explained and how Weibull probability distribution works.
The genotype and phenotype terms were created by Wilhelm Johannsen in 1911. Genotype is the complete hereditary information of an organism, even if it is not expressed. The Phenotype is a feature observed in the organism, such as morphology, development or behavior. This distinction is fundamental in the study of trait heredity and evolution. The strategy or method to map a set of genotypes to a set of phenotypes is call genotype-phenotype mapping. Genotype of an organism is an important factor in the development of its phenotype, but it is not the only one. Even two organisms with the same genotypes normally differ in their phenotypes [21].
The first studies about genotype-phenotype mapping in computation were elaborated on genetic programming in Banzhaf and Keller's work [23], they focused on Motoo Kimura's bio molecular research in 1968, where changes in molecules genotypes are phenotypically neutral are postulated, it means, several genotypes codified the same phenotype. These research is called Neutral Theory of Molecular Evolution of Motoo Kimura. Banzhaf and Keller proposed a genotype-phenotype mapping for genetic programming where they modified the non-viable genotypes to correct them with others nearby. An example in genetic algorithms is show in [25], where authors described a strategy to guide the search according with mutation. The evolutive advantage is given by some synonyms mutate towards more significant phenotypes than others.
Synonyms are explored in a more detailed way in Rothlauf y Goldberg's work [26] where they described that not all kind of representations are useful in the context of evolutive algorithms. A representation is defined as synonomically redundant if the genotypes associated with the same phenotype are similar, this concept is formally defined in terms of the sum of all distances between pairs of genotypes obtained from the set of genotypes that encodes a phenotype. If each phenotype is represented by the same number of genotypes, it means that the representation is uniformly redundant, one genotype is neighbor to another if the distance between them is the least possible, if neighboring genotypes throw neighboring phenotypes, the representation have a high locality.
In Ohnishi's work [27] is explained a mutation-based evolutionary algorithm that evolves genotypic genes for regulating developmental timing of phenotypic values. The genotype sequentially generates a given number of entire phenotypes and then finishes its life at each generation. Each genotypic gene represents a cycle time of changing probability to determine its corresponding phenotypic value in a life span of the genotype.
In Fagan's work [28] are explained some approaches that use genotype-phenotype mapping in genetic programming, and they focused on performance, it means, they examined different ways of implementing this mapping from chromosome to solution and investigate the possibility of the existence of mappings that are more suited to certain types of problems, or if a general mapping can be found which exhibits acceptable performance across all problem domains.
Probability distribution functions were used to define the mapping function in which a genotype represents a phenotype, these functions were used indirectly to generate values for each parameter in the algorithms, and this is the reason to explain each section of the proposed mechanism.
A probability distribution is a mathematical function that represent the probabilities of occurrence of different possible outcomes in an experiment, it is a description of a random phenomenon according with their respective probabilities of events.
There are a lot of probability distribution, but this research focuses on evaluate 3 different distributions to select the best one, normal, exponential and Weibull distributions were tested. The Weibull distribution function got the best results according to proposes of this research, due to it is too simple to manipulate the values of its parameters to generate many forms, and this behavior helps the adaptation of each parameter on a genetic algorithm. That's the reason Weibull distribution is the only probability distribution mentioned on this research.
Weibull distributions can be seen as a generalization of the exponential distribution as such as the gamma distribution. This distribution is commonly used for modeling reliability or survival data. It has two parameters: $ k > 0$ is the shape parameter and $\lambda > 0 $ is the scale parameter, which allow it to handle increasing, decreasing or constant failure-rates. Weibull distribution effectively describes the time we have to wait for one event to occur, if that event becomes more or less likely with time. Here the $ k$ parameter describes how quickly the probability ramps up.
It is defined as:
$
f_{x}(x) = \left\{ \right.
$
|
(11) |
Its complementary cumulative distribution function is a stretched exponential function. If the quantity x is a time-to-failure, the Weibull distribution gives a distribution where the failure rate is proportional to a time power. The pdf for the Weibull distribution drops off much more quickly (for $ k > 1$) or slowly (for $ k < 1$) than a gamma distribution. In the case where $ k = 1$, they both reduce to the exponential distribution, this particularity makes this distribution special against others. The shape parameter $ k$ can be interpreted directly as follows:
● A value of $ k < 1 $ indicates that the failure rate decreases over time. If there is significant mortality or defective items failing early, and the failure rate decrease over time as the defective items are eliminated of population.
● A value of $ k = 1$ indicates that the failure rate is constant over time. This suggest that random external events are causing mortality or failure.
● A value of $ k > 1$ indicates that the failure rate increases over time. If there is an aging behavior or parts that start to fail as time goes on.
Figure 2 shows how this distribution changes according with different values for parameters.
The algorithm is based on the idea of a genotype-phenotype mapping function which generates random values for each parameter in genetic algorithm based on a Weibull distribution, the idea is to incorporate parameters of the algorithm to evolve in each generation with the population, but, each individual has their own value for each parameter, it means, each individual has their own mutation probability and selection probability.
First, to generate each individual for genetic algorithms is needed to consider two aspects, the first one is the individual size, it represents the original size for a solution that solves the problem; the second aspect is focused on considered size for the Weibull distribution parameters, each individual on population contains the own parameters $k $ and $\lambda $ to generate a Weibull distribution (see Figure 3).
Segment on each individual that consider the Weibull distribution parameters needs satisfy two features: a) Size needs to be even. b) Size is represented in binary code. The feature (a) is defined due to segment is divided in two parts, the first one represents the value for $ \lambda $ and the second part represents the value for $ k$, and both parameters need the same quantity of binary elements to map their values in decimal integers. The feature (b) is defined because is easier to work with binary values in most of the algorithms, it is recommended that the parameters representation must be binary, to satisfy this representation is needed a mapping function that transforms binary values into integer decimal values for each parameter (see Figure 4).
To define the size of the mapping function to transform the parameters, it is necessary to define the lower and upper limits that will be used for the Weibull parameters; the minimum recommended value for $ k$ and $ \lambda $ is 1, to avoid the form that this distribution has between the 0-1 values, because it tends to be a logarithmic distribution, and it does not give good results for mechanism purposes; the upper limit value is recommended to use values according to the number of binary bits to be used, if it is necessary to handle only 3 bits it is recommended to use a limit size of 8, if it needs to use 4 bits a limit of 16, for 5 bits a limit of 32, and so on, this is to avoid normalizing from decimal to binary values (see Figure 5).
With parameters $ k$ and $ \lambda $ is possible to generate a Weibull distribution, it just needed to substitute these values on Equation 4, then the distribution generate the Cumulative Distribution Function (CDF). The CDF is used to generate random numbers in a range defined by the Weibull distribution through the Inverse Transformation method.
Inverse Transformation is a method that used the cumulative distribution function F(x) of the Weibull distribution, being that F(x) is defined in the interval (0, 1), it is possible to generate a uniform random number $ R$ and try to determine the value of the random variable for which the cumulative distribution is equal to $ R$, it means, simulated value of the random variable that follows a probability distribution f(x) is determined by solving the following equation:
$ F(x) = R \rightarrow x = F^{-1}(R) $ | (12) |
And for Weibull distribution the inverse function is:
$ F(x) = 1-e^{-(x / \lambda)^{k}} \rightarrow x = \lambda[-L n(1-u)]^{\frac{1}{k}} $ | (13) |
where ${\rm{u}} $ is a random value $\mathrm{u} \in \mathrm{U}(0, 1) $.
Finally, it is necessary to generate a mapping function that transforms the obtained random value to a value that represents a parameter of the algorithms, in GA, the parameters that the algorithm used are commonly probabilities and this is the reason that the parameters need to be mapped between 0 and 1 values.
The mapping function is defined to evaluate and to obtain the parameter values for genetic algorithm. The value obtained from the CDF is divided by the maximum value that can be obtained in the Weibull distribution, this is to normalize it in the range of 0 and 1, and it is the new value for each parameter.
Genetic algorithms have two main operators, crossover and mutation, and their values can be between 0 and 1. The idea of how the mechanism works is based on the mutation and crossover operators for binary individuals, it means, the segment that represents the parameters on each individual will evolve with this genetic operators on each generation, but the operators will focus on evolve this segment such as if algorithm were solving a binary problem; if the original problem is defined in a non-binary space, such as combinatorial problems, the segment that represents the individuals will be solved with genetic operators focused on this kind of problem, nevertheless, the segment which represent the parameters will be solved as a binary problem. The pseudo-code that represents this idea is explained in Figure A1 Appendix A.
The implementations were tested and benchmarked in 3 experiments. In the first experiment the One-Max problem was tested, this problem was configured on two instances with 50 and 500 bits length respectively. The second experiment was tested on TSP problem with TSPLib where 2 instances were obtained called "a280" and "ulysses22", where problems have 280 and 22 cities respectively, the matrix distance is obtained with Euclidean distance between cities. In third experiment 3-SAT problem was evaluated, this problem was configured on two instances with 20 variables and 91 clauses, these instances where selected from SATLib library, the files are in DIMACS CNF format, a standard format of instances of SAT developed in the nineties to express Boolean satisfaction problems in a normal conjunctive way (that is, where the problem is expressed as a conjunction of clauses and a clause is a disjunction of literals). On each dynamic instance, the changes were realized every 50 generations and every 300 generations. For both experiments, 100 runs were generated for each instance, and the average of the results were used to compare the algorithms. Due to the particular features of 3-SAT problems, 4 rules to identify dynamism in problems were: problem was updated every 50 generations and every 300 generations, and the process was to delete clauses and to add clauses. All experiments were tested in a computer with Intel i7-4800MQ processor with 2.7 GHz and 16GB RAM. Implementations were coded using Java version 8, jdk 1:8:092 and jre 1:8:0144.
The algorithms used to compare the mechanism proposed was selected by their results in an analysis realized previously through a Design of Experiments (DOE) to prove which configuration(s) of genetic algorithms are the best to solve these problems. The parameters used to generate these configurations were population, mutation probability and crossover probability. The values considered for each parameter on this DOE were:
Mutation Probability = .1, .5 y.9
Crossover Probability = .1, .5 y.9
Population = 50, 150, 300
With these values were obtained 27 different configurations for the GA. The results obtained in the first experiment for these 27 algorithms are shown in the Figures B1-B2. These results demonstrate that each configuration has a specific behavior when it tries to get the result for genetic algorithm, this is the reason that for each combination values for each parameter were created different clusters to identify the main feature that solve in a better way that problem, in case of the first experiment in its dynamic environment, results are shown in Figures B3-B4. These behaviors were similar on each problem, in Figures B5-B6 are results obtained by experiment 2 in static environment and in Figures B7-B8 in dynamic state. For the last experiment, in Figure B9-B10 are shown only results obtained for static environment to these configurations in genetic algorithm, due to the particular features described previously on this section.
The algorithms were selected according to the best results obtained and the population with a balance between exploration and exploitation strategies, all results obtained for each configuration on each problem defined in experiments were classified on different clusters, in Figures B11-B12 are shown the clusters for the analysis on One-Max problem, in Figures B13-B14 for TSP problem and in Figures B15-B16 for 3-SAT problem, on these graphs, a clustering strategy was used to identify the features of the algorithms with the best results for each problem. There exist three different clusters, the first one represents the better configurations that get best average global fitness and the population is to near of the optimal, they are represented by the biggest circles in the graphs, the features that describe these behaviors for each cluster is represented on Table 1.
Cluster | Problem | Features |
1 | One Max | Low mutation values and crossover does not have high values |
2 | One Max | All algorithms with mutation value of.5 |
3 | One Max | Mutation and crossover values are proportionally inverse |
1 | TSP | The mutation is low |
2 | TSP | The mutation has intermediate values |
3 | TSP | The mutation has the highest values |
1 | 3-SAT | The crossover has intermediate and low values, and mutation or too low or too high. |
2 | 3-SAT | Is not characterized for a special configuration |
3 | 3-SAT | Mutation has intermediate and high values |
According with clusters and features identified for each configurations to solve these problems, for each problem, the best configurations of parameters were selected, for each problem 3 different configurations such as is shown in Table 2.
Algorithm | Problem | Population Value | Cross Over Value | Mutation Value |
GA9 | One-Max | 300 | 0.9 | 0.1 |
GA18 | One-Max | 300 | 0.9 | 0.5 |
GA27 | One-Max | 300 | 0.9 | 0.9 |
GA1 | Dynamic One-Max | 50 | 0.1 | 0.1 |
GA10 | Dynamic One-Max | 50 | 0.1 | 0.5 |
GA19 | Dynamic One-Max | 50 | 0.1 | 0.9 |
GA19 | TSP | 50 | 0.1 | 0.9 |
GA20 | TSP | 150 | 0.1 | 0.9 |
GA21 | TSP | 300 | 0.1 | 0.9 |
GA11 | Dynamic TSP | 150 | 0.1 | 0.5 |
GA20 | Dynamic TSP | 150 | 0.1 | 0.9 |
GA21 | Dynamic TSP | 300 | 0.1 | 0.9 |
GA1 | 3-SAT | 50 | 0.1 | 0.1 |
GA10 | 3-SAT | 50 | 0.1 | 0.5 |
GA20 | 3-SAT | 150 | 0.1 | 0.9 |
For each problem several configurations for genetic algorithm have good results to solve it, to demonstrate these results, on Figure B17, Figure B18 and Figure B19 are shown the three configurations with best results for each problem.
The parameters to test the mechanism were mutation probability and crossover probability. The mechanism proposed was evaluated in static problems and in dynamic problems, results are shown in Figure 6 and Figure 7, to ensure that it will have good results in static problems and not only in dynamic problems, the main objective for this research is to evaluate if mechanism help genetic algorithms to change the exploration and exploitation strategies when combinatorial problems have dynamic changes over time.
In Figures 8-11 are represented the results about dynamic problems, in both cases the mechanism improves the genetic algorithm and it helps to track the optimal solution when a change occurs, in all graphics the mechanism is represented by the blue line with the best results. On each algorithm, results obtained for each configuration are shown in Table 3 for One Max, Table 4 Dynamic One Max, Table 5 TSP and Table 6 Dynamic TSP.
Algorithm | Best | Average | Mode | Minimum |
GA1 | 50 | 47.5922 | 50 | 46.64 |
GA9 | 50 | 48.2436 | 50 | 47.66 |
GA10 | 50 | 47.7099 | 50 | 47.28 |
GA11 | 50 | 45.2242 | 49 | 40.6667 |
GA18 | 50 | 48.252 | 50 | 48.0067 |
GA19 | 50 | 47.7065 | 50 | 47.6082 |
GA21 | 50 | 30.046166 | 47 | 27.57 |
GA27 | 50 | 48.2448 | 50 | 48.0633 |
GA Weibull | 50 | 32.8611 | 50 | 27.05 |
Algorithm | Best | Average | Mode | Minimum |
GA1 | 49 | 45.2046 | 49 | 42.1 |
GA9 | 50 | 36.8534 | 50 | 24.2 |
GA10 | 49 | 45.3747 | 49 | 42.87 |
GA11 | 48 | 44.0205 | 43 | 37.5267 |
GA18 | 50 | 36.8391 | 50 | 24.7533 |
GA19 | 49 | 45.3738 | 49 | 42.83 |
GA21 | 50 | 36.1885 | 50 | 22.2767 |
GA27 | 50 | 36.8146 | 50 | 24.8033 |
GA Weibull | 50 | 49.4188 | 50 | 45.64 |
Algorithm | Best | Average | Mode |
GA1 | 76.8119 | 89.5805 | 76.8119 |
GA9 | 114.0008 | 146.4622 | 126.7976 |
GA10 | 76.1759 | 88.05 | 76.1759 |
GA11 | 76.3119 | 88.0503 | 76.3119 |
GA18 | 109.7748 | 146.1850 | 163.4686 |
GA19 | 75.5088 | 87.5741 | 75.5088 |
GA21 | 75.5088 | 87.6917 | 75.5088 |
GA27 | 107.2962 | 146.1011 | 125.7882 |
GA Weibull | 75.3097 | 80.4664 | 79.1822 |
Algorithm | Best | Average | Mode |
GA1 | 63.4885 | 97.7229 | 87.438 |
GA9 | 106.0332 | 173.1659 | 144.1324 |
GA10 | 66.1179 | 95.8859 | 78.0715 |
GA11 | 66.211 | 94.5666 | 76.374 |
GA18 | 94.6573 | 173.7578 | 214.5497 |
GA19 | 61.852 | 95.5419 | 70.1903 |
GA21 | 64.335 | 95.7753 | 77.8263 |
GA27 | 102.047 | 173.8176 | 140.1949 |
GA Weibull | 61.6235 | 75.8967 | 79.2923 |
Algorithm | Best | Average | Mode | Minimum |
GA1 | 91 | 89.9336 | 91 | 89.16 |
GA10 | 91 | 89.9899 | 91 | 89.5067 |
GA20 | 91 | 88.6487 | 91 | 88.1633 |
GA Weibull | 91 | 90.6304 | 90.84 | 88.44 |
Algorithm | Best | Average | Mode | Maximum |
GA1 | 179.67 | 177.9916 | 179 | 183 |
GA10 | 180.18 | 178.4567 | 180 | 183 |
GA20 | 180.19 | 176.6554 | 180 | 183 |
GA Weibull | 108.02 | 17.7718 | 180 | 184 |
Algorithm | Best | Average | Mode | Maximum |
GA1 | 53 | 52.4398 | 53 | 53 |
GA10 | 53 | 52.4388 | 53 | 53 |
GA20 | 53 | 51.5941 | 53 | 53 |
GA Weibull | 53 | 52.8658 | 53 | 53 |
In case of results for 3-SAT problem, 4 rules were defined to identify dynamism in this kind of problems. In Figure 12 and Figure 13 are represented the best and average results for dynamic environment for 3-SAT problem adding 5 clauses each 300 generations. In Figure 14 and Figure 15 are represented the best and average results for dynamic 3-SAT problem deleting 2 clauses every 300 generations. On this experiment is represented with figures just the 3-SAT problem with changes every 300 generations, in the other case, changes every 50 generations, has the same behavior but the graphs do not show the changes as clearly as they are shown in these figures.
On each algorithm, results obtained for each configuration are shown in Table 3 for static 3-SAT, Table 4 Dynamic 3-SAT problem adding 5 clauses every 50 generations, Table 4 TSP and Table 5 Dynamic TSP. In the next tables, the information is limited to shown the three best configurations for genetic algorithm to solve this problem due to the large amount of information on each table.
Algorithm | Best | Average | Mode | Maximum |
GA1 | 105.14 | 103.9322 | 105 | 106 |
GA10 | 105.04 | 103.8655 | 105 | 106 |
GA20 | 105.13 | 105.5007 | 105 | 106 |
GA Weibull | 104.86 | 104.6524 | 105 | 106 |
Algorithm | Best | Average | Mode | Maximum |
GA1 | 75.99 | 75.1317 | 76 | 76 |
GA10 | 76 | 73.8836 | 76 | 76 |
GA20 | 76 | 75.0646 | 76 | 76 |
GA Weibull | 75.92 | 75.7808 | 76 | 76 |
The main conclusion is, at least for the proposed experiments, if genetic algorithm uses the mechanism proposed, it improves a lot the strategy to search optimal solutions in dynamic problems. When the problem changes overtime, the mechanism does not have any problem to adapt when changes occur, it has the advantage to find the optimum to fast and it helps the population to converge to this value(s). When a change occur, the algorithm is helped by the mechanism to get out of local optimum solution and to find another optimal solution, it indicates that the self-adaptation is responsible to change the diversification by an intensification strategy when the algorithm needs to find solutions in new areas, moreover, it changes diversification by intensification when there not exist changes in the solution space and the algorithm needs to find solutions in the shortest possible time, the mechanism can manage the changes between these strategies because it increased or decreased the mutation and crossover probabilities according with needs of the algorithm, due to each individual in the population has the own probabilities, each individual has different mutation and crossover probabilities and these values depend of a distribution probability function which evolves at the same time that the population evolves, it means, each individual will evolve independently depending on the individuals which it intersects and the mutations that receives on each generation, and in case of changes in the environment, the evolution of its parameters will be as drastic as possible depending on the fitness it have in the moment of changes. Furthermore, if the solution space has more than one optimal solutions, the mechanism improves the possibility to find them all.
Solutions demonstrates that mechanism improves the performance of genetic algorithms to solve these problems, and it indicates that it can be implemented in generalized problems as vehicle routing problems in some practical cases, just is needed to adjust objective function, variables and restrictions of real problem.
As future work, this mechanism will be implemented in other algorithms as Particle Swarm Optimization and Ant Colony Optimization, it will be used to solve more complex problems and to modify another parameters such as population in genetic algorithms. It will be compared with existing self-adaptive mechanism and it will be tested to solve vehicle routing problems in logistic real cases.
We would like to thank CONACyT for funding this research.
The authors declare no conflict of interest.
[1] |
Zerangue N, Kavanaugh MP (1996) Flux coupling in a neuronal glutamate transporter. Nature 383: 634-637. doi: 10.1038/383634a0
![]() |
[2] |
Danbolt NC (2001) Glutamate uptake. Prog Neurobiol 65: 1-105. doi: 10.1016/S0301-0082(00)00067-8
![]() |
[3] | Hertz L (1979) Functional interactions between neurons and astrocytes. I. Turnover and metabolism of putative amino acid transmitters. ProgNeurobiol 13: 277-323. |
[4] |
Broer S, Brookes N (2001) Transfer of glutamine between astrocytes and neurons. J Neurochem 77: 705-719. doi: 10.1046/j.1471-4159.2001.00322.x
![]() |
[5] | Drejer J, Larsson OM, Schousboe A (1982) Characterization of L-glutamate uptake into and release from astrocytes and neurons cultured from differnt brain regions. ExpBrain Res 47: 259-269. |
[6] | Schousboe A, Hertz L (1981) Role of astroglial cells in glutamate homeostasis. Adv Biochem Psychopharmacol 27: 103-113. |
[7] | Rauen T, Taylor WR, Kuhlbrodt K, et al. (1998) High-affinity glutamate transporters in the rat retina: a major role of the glial glutamate transporter GLAST-1 in transmitter clearance. Cell Tissue Res 291: 19-31. |
[8] |
Rauen T, Wiessner M (2000) Fine tuning of glutamate uptake and degradation in glial cells: common transcriptional regulation of GLAST1 and GS. Neurochem Int 37: 179-189. doi: 10.1016/S0197-0186(00)00021-8
![]() |
[9] |
Furness DN, Dehnes Y, Akhtar AQ, et al. (2008) A quantitative assessment of glutamate uptake into hippocampal synaptic terminals and astrocytes: new insights into a neuronal role for excitatory amino acid transporter 2 (EAAT2). Neuroscience 157: 80-94. doi: 10.1016/j.neuroscience.2008.08.043
![]() |
[10] |
Pines G, Danbolt NC, Bjoras M, et al. (1992) Cloning and expression of a rat brain L-glutamate transporter. Nature 360: 464-467. doi: 10.1038/360464a0
![]() |
[11] |
Storck T, Schulte S, Hofmann K, et al. (1992) Structure, expression, and functional analysis of a Na(+)-dependent glutamate/aspartate transporter from rat brain. Proc Natl Acad Sci U S A 89: 10955-10959. doi: 10.1073/pnas.89.22.10955
![]() |
[12] | Arriza JL, Fairman WA, Wadiche JI, et al. (1994) Functional comparisons of three glutamate transporter subtypes cloned from human motor cortex. J Neurosci 14: 5559-5569. |
[13] |
Tanaka K, Watase K, Manabe T, et al. (1997) Epilepsy and exacerbation of brain injury in mice lacking the glutamate transporter GLT-1. Science 276: 1699-1702. doi: 10.1126/science.276.5319.1699
![]() |
[14] |
Bjornsen LP, Hadera MG, Zhou Y, et al. (2014) The GLT-1 (EAAT2; slc1a2) glutamate transporter is essential for glutamate homeostasis in the neocortex of the mouse. J Neurochem 128: 641-649. doi: 10.1111/jnc.12509
![]() |
[15] |
Rauen T, Wiessner M, Sullivan R, et al. (2004) A new GLT1 splice variant: cloning and immunolocalization of GLT1c in the mammalian retina and brain. Neurochem Int 45: 1095-1106. doi: 10.1016/j.neuint.2004.04.006
![]() |
[16] |
Sullivan R, Rauen T, Fischer F, et al. (2004) Cloning, transport properties, and differential localization of two splice variants of GLT-1 in the rat CNS: Implications for CNS glutamate homeostasis. Glia 45: 155-169. doi: 10.1002/glia.10317
![]() |
[17] |
Lee A, Anderson AR, Beasley SJ, et al. (2012) A new splice variant of the glutamate-aspartate transporter: cloning and immunolocalization of GLAST1c in rat, pig and human brains. J Chem Neuroanat 43: 52-63. doi: 10.1016/j.jchemneu.2011.10.005
![]() |
[18] |
Grewer C, Gameiro A, Rauen T (2014) SLC1 glutamate transporters. Pflugers Arch 466: 3-24. doi: 10.1007/s00424-013-1397-7
![]() |
[19] |
Rauen T (2000) Diversity of glutamate transporter expression and function in the mammalian retina. Amino Acids 19: 53-62. doi: 10.1007/s007260070033
![]() |
[20] |
Rauen T, Kanner BI (1994) Localization of the glutamate transporter GLT-1 in rat and macaque monkey retinae. Neurosci Lett 169: 137-140. doi: 10.1016/0304-3940(94)90375-1
![]() |
[21] |
Wiessner M, Fletcher EL, Fischer F, et al. (2002) Localization and possible function of the glutamate transporter, EAAC1, in the rat retina. Cell Tissue Res 310: 31-40. doi: 10.1007/s00441-002-0612-1
![]() |
[22] |
Holmseth S, Dehnes Y, Huang YH, et al. (2012) The density of EAAC1 (EAAT3) glutamate transporters expressed by neurons in the mammalian CNS. J Neurosci 32: 6000-6013. doi: 10.1523/JNEUROSCI.5347-11.2012
![]() |
[23] | Dehnes Y, Chaudhry FA, Ullensvang K, et al. (1998) The glutamate transporter EAAT4 in rat cerebellar Purkinje cells: a glutamate-gated chloride channel concentrated near the synapse in parts of the dendritic membrane facing astroglia. J Neurosci 18: 3606-3619. |
[24] |
Mim C, Balani P, Rauen T, et al. (2005) The Glutamate Transporter Subtypes EAAT4 and EAATs 1-3 Transport Glutamate with Dramatically Different Kinetics and Voltage Dependence but Share a Common Uptake Mechanism. J Gen Physiol 126: 571-589. doi: 10.1085/jgp.200509365
![]() |
[25] |
Gincel D, Regan MR, Jin L, et al. (2007) Analysis of cerebellar Purkinje cells using EAAT4 glutamate transporter promoter reporter in mice generated via bacterial artificial chromosome-mediated transgenesis. Exp Neurol 203: 205-212. doi: 10.1016/j.expneurol.2006.08.016
![]() |
[26] |
Kovermann P, Machtens JP, Ewers D, et al. (2010) A conserved aspartate determines pore properties of anion channels associated with excitatory amino acid transporter 4 (EAAT4). J Biol Chem 285: 23676-23686. doi: 10.1074/jbc.M110.126557
![]() |
[27] |
Arriza JL, Eliasof S, Kavanaugh MP, et al. (1997) Excitatory amino acid transporter 5, a retinal glutamate transporter coupled to a chloride conductance. Proc Natl Acad Sci U S A 94: 4155-4160. doi: 10.1073/pnas.94.8.4155
![]() |
[28] |
Wersinger E, Schwab Y, Sahel JA, et al. (2006) The glutamate transporter EAAT5 works as a presynaptic receptor in mouse rod bipolar cells. J Physiol 577: 221-234. doi: 10.1113/jphysiol.2006.118281
![]() |
[29] |
Gameiro A, Braams S, Rauen T, et al. (2011) The Discovery of Slowness: Low-Capacity Transport and Slow Anion Channel Gating by the Glutamate Transporter EAAT5. Biophysical journal 100: 2623-2632. doi: 10.1016/j.bpj.2011.04.034
![]() |
[30] | Hediger MA, Kanai Y, You G, et al. (1995) Mammalian ion-coupled solute transporters. JPhysiolLond 482: 7S-17S. |
[31] |
Bailey CG, Ryan RM, Thoeng AD, et al. (2011) Loss-of-function mutations in the glutamate transporter SLC1A1 cause human dicarboxylic aminoaciduria. J Clin Invest 121: 446-453. doi: 10.1172/JCI44474
![]() |
[32] |
Duerson K, Woltjer RL, Mookherjee P, et al. (2009) Detergent-insoluble EAAC1/EAAT3 aberrantly accumulates in hippocampal neurons of Alzheimer's disease patients. Brain Pathol 19: 267-278. doi: 10.1111/j.1750-3639.2008.00186.x
![]() |
[33] |
Revett TJ, Baker GB, Jhamandas J, et al. (2013) Glutamate system, amyloid ss peptides and tau protein: functional interrelationships and relevance to Alzheimer disease pathology. J Psychiatry Neurosci 38: 6-23. doi: 10.1503/jpn.110190
![]() |
[34] | Rothstein JD (2009) Current hypotheses for the underlying biology of amyotrophic lateral sclerosis. Ann Neurol 65 Suppl 1: S3-9. |
[35] |
Lang UE, Borgwardt S (2013) Molecular Mechanisms of Depression: Perspectives on New Treatment Strategies. Cell Physiol Biochem 31: 761-777. doi: 10.1159/000350094
![]() |
[36] |
Crino PB, Jin H, Shumate MD, et al. (2002) Increased expression of the neuronal glutamate transporter (EAAT3/EAAC1) in hippocampal and neocortical epilepsy. Epilepsia 43: 211-218. doi: 10.1046/j.1528-1157.2002.35001.x
![]() |
[37] |
Estrada-Sanchez AM, Rebec GV (2012) Corticostriatal dysfunction and glutamate transporter 1 (GLT1) in Huntington's disease: interactions between neurons and astrocytes. Basal Ganglia 2: 57-66. doi: 10.1016/j.baga.2012.04.029
![]() |
[38] | Rao VL, Dogan A, Todd KG, et al. (2001) Antisense knockdown of the glial glutamate transporter GLT-1, but not the neuronal glutamate transporter EAAC1, exacerbates transient focal cerebral ischemia-induced neuronal damage in rat brain. J Neurosci 21: 1876-1883. |
[39] |
Grewer C, Gameiro A, Zhang Z, et al. (2008) Glutamate forward and reverse transport: from molecular mechanism to transporter-mediated release after ischemia. IUBMB Life 60: 609-619. doi: 10.1002/iub.98
![]() |
[40] |
Ketheeswaranathan P, Turner NA, Spary EJ, et al. (2011) Changes in glutamate transporter expression in mouse forebrain areas following focal ischemia. Brain Res 1418: 93-103. doi: 10.1016/j.brainres.2011.08.029
![]() |
[41] |
Seki Y, Feustel PJ, Keller RW, et al. (1999) Inhibition of ischemia-induced glutamate release in rat striatum by dihydrokinate and an anion channel blocker. Stroke 30: 433-440. doi: 10.1161/01.STR.30.2.433
![]() |
[42] | Azami Tameh A, Clarner T, Beyer C, et al. (2013) Regional regulation of glutamate signaling during cuprizone-induced demyelination in the brain. Ann Anat. |
[43] |
Karlsson RM, Tanaka K, Heilig M, et al. (2008) Loss of glial glutamate and aspartate transporter (excitatory amino acid transporter 1) causes locomotor hyperactivity and exaggerated responses to psychotomimetics: rescue by haloperidol and metabotropic glutamate 2/3 agonist. Biol Psychiatry 64: 810-814. doi: 10.1016/j.biopsych.2008.05.001
![]() |
[44] |
Karlsson RM, Tanaka K, Saksida LM, et al. (2009) Assessment of glutamate transporter GLAST (EAAT1)-deficient mice for phenotypes relevant to the negative and executive/cognitive symptoms of schizophrenia. Neuropsychopharmacology 34: 1578-1589. doi: 10.1038/npp.2008.215
![]() |
[45] |
Adamczyk A, Gause CD, Sattler R, et al. (2011) Genetic and functional studies of a missense variant in a glutamate transporter, SLC1A3, in Tourette syndrome. Psychiatr Genet 21: 90-97. doi: 10.1097/YPG.0b013e328341a307
![]() |
[46] |
Reyes N, Ginter C, Boudker O (2009) Transport mechanism of a bacterial homologue of glutamate transporters. Nature 462: 880-885. doi: 10.1038/nature08616
![]() |
[47] |
Verdon G, Boudker O (2012) Crystal structure of an asymmetric trimer of a bacterial glutamate transporter homolog. Nat Struct Mol Biol 19: 355-357. doi: 10.1038/nsmb.2233
![]() |
[48] |
Yernool D, Boudker O, Jin Y, et al. (2004) Structure of a glutamate transporter homologue from Pyrococcus horikoshii. Nature 431: 811-818. doi: 10.1038/nature03018
![]() |
[49] |
Jardetzky O (1966) Simple allosteric model for membrane pumps. Nature 211: 969-970. doi: 10.1038/211969a0
![]() |
[50] |
Owe SG, Marcaggi P, Attwell D (2006) The ionic stoichiometry of the GLAST glutamate transporter in salamander retinal glia. J Physiol 577: 591-599. doi: 10.1113/jphysiol.2006.116830
![]() |
[51] |
Kanai Y, Nussberger S, Romero MF, et al. (1995) Electrogenic properties of the epithelial and neuronal high affinity glutamate transporter. J Biol Chem 270: 16561-16568. doi: 10.1074/jbc.270.28.16561
![]() |
[52] | Wadiche JI, Kavanaugh MP (1998) Macroscopic and microscopic properties of a cloned glutamate transporter/chloride channel. J Neurosci 18: 7650-7661. |
[53] | Otis TS, Kavanaugh MP (2000) Isolation of current components and partial reaction cycles in the glial glutamate transporter EAAT2. J Neurosci 20: 2749-2757. |
[54] | Otis TS, Jahr CE (1998) Anion currents and predicted glutamate flux through a neuronal glutamate transporter. J Neurosci 18: 7099-7110. |
[55] | Bergles DE, Tzingounis AV, Jahr CE (2002) Comparison of coupled and uncoupled currents during glutamate uptake by GLT-1 transporters. J Neurosci 22: 10153-10162. |
[56] |
Grewer C, Watzke N, Wiessner M, et al. (2000) Glutamate translocation of the neuronal glutamate transporter EAAC1 occurs within milliseconds. Proc Natl Acad Sci U S A 97: 9706-9711. doi: 10.1073/pnas.160170397
![]() |
[57] |
Watzke N, Bamberg E, Grewer C (2001) Early intermediates in the transport cycle of the neuronal excitatory amino acid carrier EAAC1. J Gen Physiol 117: 547-562. doi: 10.1085/jgp.117.6.547
![]() |
[58] | Mwaura J, Tao Z, James H, et al. (2012) Protonation state of a conserved acidic amino acid involved in Na(+) binding to the glutamate transporter EAAC1. ACS Chem Neurosci 12: 1073-1083. |
[59] | Diamond JS, Jahr CE (1997) Transporters buffer synaptically released glutamate on a submillisecond time scale. J Neurosci 17: 4672-4687. |
[60] |
Mim C, Tao Z, Grewer C (2007) Two conformational changes are associated with glutamate translocation by the glutamate transporter EAAC1. Biochemistry 46: 9007-9018. doi: 10.1021/bi7005465
![]() |
[61] |
Wadiche JI, Arriza JL, Amara SG, et al. (1995) Kinetics of a human glutamate transporter. Neuron 14: 1019-1027. doi: 10.1016/0896-6273(95)90340-2
![]() |
[62] |
Loo DD, Hazama A, Supplisson S, et al. (1993) Relaxation kinetics of the Na+/glucose cotransporter. Proc Natl Acad Sci U S A 90: 5767-5771. doi: 10.1073/pnas.90.12.5767
![]() |
[63] | Lu CC, Hilgemann DW (1999) GAT1 (GABA:Na+:Cl-) cotransport function. Kinetic studies in giant Xenopus oocyte membrane patches. J Gen Physiol 114: 445-457. |
[64] |
Grewer C, Zhang Z, Mwaura J, et al. (2012) Charge compensation mechanism of a Na+-coupled, secondary active glutamate transporter. J Biol Chem 287: 26921-26931. doi: 10.1074/jbc.M112.364059
![]() |
[65] |
Zhang Z, Tao Z, Gameiro A, et al. (2007) Transport direction determines the kinetics of substrate transport by the glutamate transporter EAAC1. Proc Natl Acad Sci U S A 104: 18025-18030. doi: 10.1073/pnas.0704570104
![]() |
[66] |
Wadiche JI, Amara SG, Kavanaugh MP (1995) Ion fluxes associated with excitatory amino acid transport. Neuron 15: 721-728. doi: 10.1016/0896-6273(95)90159-0
![]() |
[67] |
Eliasof S, Jahr CE (1996) Retinal glial cell glutamate transporter is coupled to an anionic conductance. Proc Natl Acad Sci U S A 93: 4153-4158. doi: 10.1073/pnas.93.9.4153
![]() |
[68] | Billups B, Rossi D, Attwell D (1996) Anion conductance behavior of the glutamate uptake carrier in salamander retinal glial cells. J Neurosci 16: 6722-6731. |
[69] |
Fairman WA, Vandenberg RJ, Arriza JL, et al. (1995) An excitatory amino-acid transporter with properties of a ligand-gated chloride channel. Nature 375: 599-603. doi: 10.1038/375599a0
![]() |
[70] |
Larsson HP, Picaud SA, Werblin FS, et al. (1996) Noise analysis of the glutamate-activated current in photoreceptors. Biophysl J 70: 733-742. doi: 10.1016/S0006-3495(96)79613-3
![]() |
[71] |
Melzer N, Biela A, Fahlke C (2003) Glutamate modifies ion conduction and voltage-dependent gating of excitatory amino acid transporter-associated anion channels. J Biol Chem 278: 50112-50119. doi: 10.1074/jbc.M307990200
![]() |
[72] | Picaud SA, Larsson HP, Grant GB, et al. (1995) Glutamate-gated chloride channel with glutamate-transporter-like properties in cone photoreceptors of the tiger salamander. J Neurophys 74: 1760-1771. |
[73] |
Watzke N, Grewer C (2001) The anion conductance of the glutamate transporter EAAC1 depends on the direction of glutamate transport. FEBS Lett 503: 121-125. doi: 10.1016/S0014-5793(01)02715-6
![]() |
[74] |
Tao Z, Grewer C (2007) Cooperation of the conserved aspartate 439 and bound amino acid substrate is important for high-affinity Na+ binding to the glutamate transporter EAAC1. J Gen Physiol 129: 331-344. doi: 10.1085/jgp.200609678
![]() |
[75] |
Boudker O, Ryan RM, Yernool D, et al. (2007) Coupling substrate and ion binding to extracellular gate of a sodium-dependent aspartate transporter. Nature 445: 387-393. doi: 10.1038/nature05455
![]() |
[76] |
Cater RJ, Vandenberg RJ, Ryan RM (2014) The domain interface of the human glutamate transporter EAAT1 mediates chloride permeation. Biophys J 107: 621-629. doi: 10.1016/j.bpj.2014.05.046
![]() |
[77] |
Huang Z, Tajkhorshid E (2008) Dynamics of the extracellular gate and ion-substrate coupling in the glutamate transporter. Biophys J 95: 2292-2300. doi: 10.1529/biophysj.108.133421
![]() |
[78] |
Shrivastava IH, Jiang J, Amara SG, et al. (2008) Time-resolved mechanism of extracellular gate opening and substrate binding in a glutamate transporter. J Biol Chem 283: 28680-28690. doi: 10.1074/jbc.M800889200
![]() |
[79] |
Huang Z, Tajkhorshid E (2010) Identification of the third Na+ site and the sequence of extracellular binding events in the glutamate transporter. Biophys J 99: 1416-1425. doi: 10.1016/j.bpj.2010.06.052
![]() |
[80] |
Bastug T, Heinzelmann G, Kuyucak S, et al. (2012) Position of the third Na+ site in the aspartate transporter GltPh and the human glutamate transporter, EAAT1. PLoS One 7: e33058. doi: 10.1371/journal.pone.0033058
![]() |
[81] |
Groeneveld M, Slotboom DJ (2010) Na(+):aspartate coupling stoichiometry in the glutamate transporter homologue Glt(Ph). Biochemistry 49: 3511-3513. doi: 10.1021/bi100430s
![]() |
[82] |
Larsson HP, Wang X, Lev B, et al. (2010) Evidence for a third sodium-binding site in glutamate transporters suggests an ion/substrate coupling model. Proc Natl Acad Sci U S A 107: 13912-13917. doi: 10.1073/pnas.1006289107
![]() |
[83] |
DeChancie J, Shrivastava IH, Bahar I (2011) The mechanism of substrate release by the aspartate transporter GltPh: insights from simulations. Mol Biosyst 7: 832-842. doi: 10.1039/C0MB00175A
![]() |
[84] |
Zomot E, Bahar I (2013) Intracellular gating in an inward-facing state of aspartate transporter Glt(Ph) is regulated by the movements of the helical hairpin HP2. J Biol Chem 288: 8231-8237. doi: 10.1074/jbc.M112.438432
![]() |
[85] |
Heinzelmann G, Kuyucak S (2014) Molecular dynamics simulations of the mammalian glutamate transporter EAAT3. PLoS One 9: e92089. doi: 10.1371/journal.pone.0092089
![]() |
[86] |
Jiang J, Shrivastava IH, Watts SD, et al. (2011) Large collective motions regulate the functional properties of glutamate transporter trimers. Proc Natl Acad Sci U S A 108: 15141-15146. doi: 10.1073/pnas.1112216108
![]() |
[87] |
Lezon TR, Bahar I (2012) Constraints imposed by the membrane selectively guide the alternating access dynamics of the glutamate transporter GltPh. Biophys J 102: 1331-1340. doi: 10.1016/j.bpj.2012.02.028
![]() |
[88] |
Das A, Gur M, Cheng MH, et al. (2014) Exploring the conformational transitions of biomolecular systems using a simple two-state anisotropic network model. PLoS Comput Biol 10: e1003521. doi: 10.1371/journal.pcbi.1003521
![]() |
[89] |
Stolzenberg S, Khelashvili G, Weinstein H (2012) Structural intermediates in a model of the substrate translocation path of the bacterial glutamate transporter homologue GltPh. J Phys Chem B 116: 5372-5383. doi: 10.1021/jp301726s
![]() |
[90] |
Grewer C, Watzke N, Rauen T, et al. (2003) Is the glutamate residue Glu-373 the proton acceptor of the excitatory amino acid carrier 1? J Biol Chem 278: 2585-2592. doi: 10.1074/jbc.M207956200
![]() |
[91] |
Heinzelmann G, Kuyucak S (2014) Molecular Dynamics Simulations Elucidate the Mechanism of Proton Transport in the Glutamate Transporter EAAT3. Biophys J 106: 2675-2683. doi: 10.1016/j.bpj.2014.05.010
![]() |
[92] |
Grewer C, Jager J, Carpenter BK, et al. (2000) A new photolabile precursor of glycine with improved properties: A tool for chemical kinetic investigations of the glycine receptor. Biochemistry 39: 2063-2070. doi: 10.1021/bi9919652
![]() |
[93] |
Grewer C, Rauen T (2005) Electrogenic glutamate transporters in the CNS: molecular mechanism, pre-steady-state kinetics, and their impact on synaptic signaling. J Membr Biol 203: 1-20. doi: 10.1007/s00232-004-0731-6
![]() |
[94] |
Gegelashvili G, Robinson MB, Trotti D, et al. (2001) Regulation of glutamate transporters in health and disease. Prog Brain Res 132: 267-286. doi: 10.1016/S0079-6123(01)32082-4
![]() |
[95] |
Santos SD, Carvalho AL, Caldeira MV, et al. (2009) Regulation of AMPA receptors and synaptic plasticity. Neuroscience 158: 105-125. doi: 10.1016/j.neuroscience.2008.02.037
![]() |
[96] |
Stephenson FA, Cousins SL, Kenny AV (2008) Assembly and forward trafficking of NMDA receptors (Review). Mol Membr Biol 25: 311-320. doi: 10.1080/09687680801971367
![]() |
[97] | Robinson MB (2002) Regulated trafficking of neurotransmitter transporters: common notes but different melodies. J Neurochem 80: 1-11. |
[98] |
Gonzalez MI, Robinson MB (2004) Protein KINASE C-Dependent Remodeling of Glutamate Transporter Function. Mol Intervent 4: 48-58. doi: 10.1124/mi.4.1.48
![]() |
[99] |
Sheldon AL, Robinson MB (2007) The role of glutamate transporters in neurodegenerative diseases and potential opportunities for intervention. Neurochem Int 51: 333-355. doi: 10.1016/j.neuint.2007.03.012
![]() |
[100] | Beart PM, O'Shea RD (2007) Transporters for L-glutamate: an update on their molecular pharmacology and pathological involvement. Br J Pharmacol 150: 5-17. |
[101] |
Poitry-Yamate CL, Vutskits L, Rauen T (2002) Neuronal-induced and glutamate-dependent activation of glial glutamate transporter function. J Neurochem 82: 987-997. doi: 10.1046/j.1471-4159.2002.01075.x
![]() |
[102] |
Benediktsson AM, Marrs GS, Tu JC, et al. (2012) Neuronal activity regulates glutamate transporter dynamics in developing astrocytes. Glia 60: 175-188. doi: 10.1002/glia.21249
![]() |
[103] |
Gonzalez-Gonzalez IM, Garcia-Tardon N, Gimenez C, et al. (2008) PKC-dependent endocytosis of the GLT1 glutamate transporter depends on ubiquitylation of lysines located in a C-terminal cluster. Glia 56: 963-974. doi: 10.1002/glia.20670
![]() |
[104] |
Sheldon AL, Gonzalez MI, Krizman-Genda EN, et al. (2008) Ubiquitination-mediated internalization and degradation of the astroglial glutamate transporter, GLT-1. Neurochem Int 53: 296-308. doi: 10.1016/j.neuint.2008.07.010
![]() |
[105] |
Martinez-Villarreal J, Garcia Tardon N, Ibanez I, et al. (2012) Cell surface turnover of the glutamate transporter GLT-1 is mediated by ubiquitination/deubiquitination. Glia 60: 1356-1365. doi: 10.1002/glia.22354
![]() |
[106] |
Sheldon AL, Gonzalez MI, Robinson MB (2006) A carboxyl-terminal determinant of the neuronal glutamate transporter, EAAC1, is required for platelet-derived growth factor-dependent trafficking. J Biol Chem 281: 4876-4886. doi: 10.1074/jbc.M504983200
![]() |
[107] |
Garcia-Tardon N, Gonzalez-Gonzalez IM, Martinez-Villarreal J, et al. (2012) Protein kinase C (PKC)-promoted endocytosis of glutamate transporter GLT-1 requires ubiquitin ligase Nedd4-2-dependent ubiquitination but not phosphorylation. J Biol Chem 287: 19177-19187. doi: 10.1074/jbc.M112.355909
![]() |
[108] |
A DA, Soragna A, Di Cairano E, et al. (2010) The Surface Density of the Glutamate Transporter EAAC1 is Controlled by Interactions with PDZK1 and AP2 Adaptor Complexes. Traffic 11: 1455-1470. doi: 10.1111/j.1600-0854.2010.01110.x
![]() |
[109] |
Traub LM (2009) Tickets to ride: selecting cargo for clathrin-regulated internalization. Nat Rev Mol Cell Biol 10: 583-596. doi: 10.1038/nrm2751
![]() |
[110] |
Sato K, Otsu W, Otsuka Y, et al. (2013) Modulatory roles of NHERF1 and NHERF2 in cell surface expression of the glutamate transporter GLAST. Biochem Biophys Res Commun 430: 839-845. doi: 10.1016/j.bbrc.2012.11.059
![]() |
[111] | Shouffani A, Kanner BI (1990) Cholesterol is required for the reconstruction of the sodium- and chloride-coupled, gamma-aminobutyric acid transporter from rat brain. J Biol Chem 265: 6002-6008. |
[112] |
Butchbach ME, Guo H, Lin CL (2003) Methyl-beta-cyclodextrin but not retinoic acid reduces EAAT3-mediated glutamate uptake and increases GTRAP3-18 expression. J Neurochem 84: 891-894. doi: 10.1046/j.1471-4159.2003.01588.x
![]() |
[113] |
Simons K, Gerl MJ (2010) Revitalizing membrane rafts: new tools and insights. Nat Rev Mol Cell Biol 11: 688-699. doi: 10.1038/nrm2977
![]() |
[114] |
Butchbach ME, Tian G, Guo H, et al. (2004) Association of excitatory amino acid transporters, especially EAAT2, with cholesterol-rich lipid raft microdomains: importance for excitatory amino acid transporter localization and function. J Biol Chem 279: 34388-34396. doi: 10.1074/jbc.M403938200
![]() |
[115] |
Zschocke J, Bayatti N, Behl C (2005) Caveolin and GLT-1 gene expression is reciprocally regulated in primary astrocytes: association of GLT-1 with non-caveolar lipid rafts. Glia 49: 275-287. doi: 10.1002/glia.20116
![]() |
[116] |
Gonzalez MI, Krizman-Genda E, Robinson MB (2007) Caveolin-1 regulates the delivery and endocytosis of the glutamate transporter, excitatory amino acid carrier 1. J Biol Chem 282: 29855-29865. doi: 10.1074/jbc.M704738200
![]() |
[117] | Ledesma MD, Dotti CG (2005) The conflicting role of brain cholesterol in Alzheimer's disease: lessons from the brain plasminogen system. Biochem Soc Symp: 129-138. |
[118] |
Tian G, Kong Q, Lai L, et al. (2010) Increased expression of cholesterol 24S-hydroxylase results in disruption of glial glutamate transporter EAAT2 association with lipid rafts: a potential role in Alzheimer's disease. J Neurochem 113: 978-989. doi: 10.1111/j.1471-4159.2010.06661.x
![]() |
[119] |
Arriza JL, Eliasof S, Kavanaugh MP, et al. (1997) Excitatory amino acid transporter 5, a retinal glutamate transporter coupled to a chloride conductance. Proc Natl Acad Sci U S A 94: 4155-4160. doi: 10.1073/pnas.94.8.4155
![]() |
[120] | Arriza JL, Fairman WA, Wadiche JI, et al. (1994) Functional comparisons of three glutamate transporter subtypes cloned from human motor cortex. J Neurosci 14: 5559-5569. |
[121] | Bridges RJ, Stanley MS, Anderson MW, et al. (1991) Conformationally defined neurotransmitter analogues. Selective inhibition of glutamate uptake by one pyrrolidine-2,4-dicarboxylate diastereomer. J Med Chem 34: 717-725. |
[122] |
Griffiths R, Dunlop J, Gorman A, et al. (1994) L-Trans-Pyrrolidine-2,4-Dicarboxylate and Cis-1-Aminocyclobutane-1,3-Dicarboxylate Behave as Transportable, Competitive Inhibitors of the High-Affinity Glutamate Transporters. Biochem Pharmacol 47: 267-274. doi: 10.1016/0006-2952(94)90016-7
![]() |
[123] | Vandenberg RJ, Mitrovic AD, Chebib M, et al. (1997) Contrasting modes of action of methylglutamate derivatives on the excitatory amino acid transporters, EAAT1 and EAAT2. Mol Pharmacol 51: 809-815. |
[124] |
Huang S, Ryan RM, Vandenberg RJ (2009) The role of cation binding in determining substrate selectivity of glutamate transporters. J Biol Chem 284: 4510-4515. doi: 10.1074/jbc.M808495200
![]() |
[125] |
Eliasof S, McIlvain HB, Petroski RE, et al. (2001) Pharmacological characterization of threo-3-methylglutamic acid with excitatory amino acid transporters in native and recombinant systems. J Neurochem 77: 550-557. doi: 10.1046/j.1471-4159.2001.00253.x
![]() |
[126] |
Kanai Y, Hediger MA (1992) Primary structure and functional characterization of a high-affinity glutamate transporter. Nature 360: 467-471. doi: 10.1038/360467a0
![]() |
[127] |
Rauen T, Jeserich G, Danbolt NC, et al. (1992) Comparative analysis of sodium-dependent L-glutamate transport of synaptosomal and astroglial membrane vesicles from mouse cortex. FEBS Lett 312: 15-20. doi: 10.1016/0014-5793(92)81401-7
![]() |
[128] | Zerangue N, Kavanaugh MP (1996) Interaction of L-cysteine with a human excitatory amino acid transporter. J Physiol 493 ( Pt 2): 419-423. |
[129] |
Roberts PJ, Watkins JC (1975) Structural requirements for the inhibition for L-glutamate uptake by glia and nerve endings. Brain Res 85: 120-125. doi: 10.1016/0006-8993(75)91016-1
![]() |
[130] | Wilson DF, Pastuszko A (1986) Transport of Cysteate by Synaptosomes Isolated from Rat-Brain - Evidence That It Utilizes the Same Transporter as Aspartate, Glutamate, and Cysteine Sulfinate. J Neurochem 47: 1091-1097. |
[131] |
Vandenberg RJ, Mitrovic AD, Johnston GAR (1998) Serine-O-sulphate transport by the human glutamate transporter, EAAT2. Br J Pharmacol 123: 1593-1600. doi: 10.1038/sj.bjp.0701776
![]() |
[132] |
Bender AS, Woodbury DM, White HS (1989) Beta-Dl-Methylene-Aspartate, an Inhibitor of Aspartate-Aminotransferase, Potently Inhibits L-Glutamate Uptake into Astrocytes. Neurochem Res 14: 641-646. doi: 10.1007/BF00964873
![]() |
[133] |
Mitrovic AD, Amara SG, Johnston GA, et al. (1998) Identification of functional domains of the human glutamate transporters EAAT1 and EAAT2. J Biol Chem 273: 14698-14706. doi: 10.1074/jbc.273.24.14698
![]() |
[134] |
Vandenberg RJ, Mitrovic AD, Johnston GA (1998) Serine-O-sulphate transport by the human glutamate transporter, EAAT2. Br J Pharmacol 123: 1593-1600. doi: 10.1038/sj.bjp.0701776
![]() |
[135] | Campiani G, De Angelis M, Armaroli S, et al. (2001) A rational approach to the design of selective substrates and potent nontransportable inhibitors of the excitatory amino acid transporter EAAC1 (EAAT3). New glutamate and aspartate analogues as potential neuroprotective agents. J Med Chem 44: 2507-2510. |
[136] |
Danbolt NC (2001) Glutamate uptake. Prog Neurobiol 65: 1-105. doi: 10.1016/S0301-0082(00)00067-8
![]() |
[137] |
Wang GJ, Chung HJ, Schnuer J, et al. (1998) Dihydrokainate-sensitive neuronal glutamate transport is required for protection of rat cortical neurons in culture against synaptically released glutamate. Eur J Neurosci 10: 2523-2531. doi: 10.1046/j.1460-9568.1998.00256.x
![]() |
[138] | Shimamoto K, Lebrun B, Yasuda-Kamatani Y, et al. (1998) DL-threo-beta-benzyloxyaspartate, a potent blocker of excitatory amino acid transporters. Mol Pharmacol 53: 195-201. |
[139] |
Boudker O, Verdon G (2010) Structural perspectives on secondary active transporters. Trends Pharmacol Sci 31: 418-426. doi: 10.1016/j.tips.2010.06.004
![]() |
[140] | Shigeri Y, Shimamoto K, Yasuda-Kamatani Y, et al. (2001) Effects of threo-beta-hydroxyaspartate derivatives on excitatory amino acid transporters (EAAT4 and EAAT5). J Neurochem 79: 297-302. |
[141] |
Shimamoto K, Shigeri Y, Yasuda-Kamatani Y, et al. (2000) Syntheses of optically pure beta-hydroxyaspartate derivatives as glutamate transporter blockers. Bioorg Med Chem Lett 10: 2407-2410. doi: 10.1016/S0960-894X(00)00487-X
![]() |
[142] |
Lebrun B, Sakaitani M, Shimamoto K, et al. (1997) New beta-hydroxyaspartate derivatives are competitive blockers for the bovine glutamate/aspartate transporter. J Biol Chem 272: 20336-20339. doi: 10.1074/jbc.272.33.20336
![]() |
[143] |
Shimamoto K, Sakai R, Takaoka K, et al. (2004) Characterization of novel L-threo-beta-benzyloxyaspartate derivatives, potent blockers of the glutamate transporters. Mol Pharmacol 65: 1008-1015. doi: 10.1124/mol.65.4.1008
![]() |
[144] | Shimamoto K, Otsubo Y, Shigeri Y, et al. (2007) Characterization of the tritium-labeled analog of L-threo-beta-benzyloxyaspartate binding to glutamate transporters. Mol Pharmacol 71: 294-302. |
[145] |
Martinov V, Dehnes Y, Holmseth S, et al. (2014) A novel glutamate transporter blocker, LL-TBOA, attenuates ischaemic injury in the isolated, perfused rat heart despite low transporter levels. Eur J Cardiothorac Surg 45: 710-716. doi: 10.1093/ejcts/ezt487
![]() |
[146] | Dunlop J, Eliasof S, Stack G, et al. (2003) WAY-855 (3-amino-tricyclo[2.2.1.02.6]heptane-1,3-dicarboxylic acid): a novel, EAAT2-preferring, nonsubstrate inhibitor of high-affinity glutamate uptake. Br J Pharmacol 140: 839-846. |
[147] |
Dunlop J, McIlvain HB, Carrick TA, et al. (2005) Characterization of novel aryl-ether, biaryl, and fluorene aspartic acid and diaminopropionic acid analogs as potent inhibitors of the high-affinity glutamate transporter EAAT2. Mol Pharmacol 68: 974-982. doi: 10.1124/mol.105.012005
![]() |
[148] |
Campiani G, Fattorusso C, De Angelis M, et al. (2003) Neuronal high-affinity sodium-dependent glutamate transporters (EAATs): targets for the development of novel therapeutics against neurodegenerative diseases. Curr Pharm Des 9: 599-625. doi: 10.2174/1381612033391261
![]() |
[149] | Funicello M, Conti P, De Amici M, et al. (2004) Dissociation of [3H]L-glutamate uptake from L-glutamate-induced [3H]D-aspartate release by 3-hydroxy-4,5,6,6a-tetrahydro-3aH-pyrrolo[3,4-d]isoxazole-4-carboxylic acid and 3-hydroxy-4,5,6,6a-tetrahydro-3aH-pyrrolo[3,4-d]isoxazole-6-carboxylic acid, two conformationally constrained aspartate and glutamate analogs. Mol Pharmacol 66: 522-529. |
[150] |
Callender R, Gameiro A, Pinto A, et al. (2012) Mechanism of inhibition of the glutamate transporter EAAC1 by the conformationally constrained glutamate analogue (+)-HIP-B. Biochemistry 51: 5486-5495. doi: 10.1021/bi3006048
![]() |
[151] |
Erichsen MN, Huynh TH, Abrahamsen B, et al. (2010) Structure-activity relationship study of first selective inhibitor of excitatory amino acid transporter subtype 1: 2-Amino-4-(4-methoxyphenyl)-7-(naphthalen-1-yl)-5-oxo-5,6,7,8-tetrahydro-4H-chromene-3-carbonitrile (UCPH-101). J Med Chem 53: 7180-7191. doi: 10.1021/jm1009154
![]() |
[152] |
Huynh THV, Shim I, Bohr H, et al. (2012) Structure-Activity Relationship Study of Selective Excitatory Amino Acid Transporter Subtype 1 (EAAT1) Inhibitor 2-Amino-4-(4-methoxyphenyl)-7-(naphthalen-1-yl)-5-oxo-5,6,7,8-tetrahydro-4H-chromene-3-carbonitrile (UCPH-101) and Absolute Configurational Assignment Using Infrared and Vibrational Circular Dichroism Spectroscopy in Combination with ab Initio Hartree-Fock Calculations. J Med Chem 55: 5403-5412. doi: 10.1021/jm300345z
![]() |
[153] |
Abrahamsen B, Schneider N, Erichsen MN, et al. (2013) Allosteric Modulation of an Excitatory Amino Acid Transporter: The Subtype-Selective Inhibitor UCPH-101 Exerts Sustained Inhibition of EAAT1 through an Intramonomeric Site in the Trimerization Domain. J Neurosci 33: 1068-1087. doi: 10.1523/JNEUROSCI.3396-12.2013
![]() |
[154] |
Rothstein JD, Patel S, Regan MR, et al. (2005) Beta-lactam antibiotics offer neuroprotection by increasing glutamate transporter expression. Nature 433: 73-77. doi: 10.1038/nature03180
![]() |
[155] |
Fontana AC, de Oliveira Beleboni R, Wojewodzic MW, et al. (2007) Enhancing glutamate transport: mechanism of action of Parawixin1, a neuroprotective compound from Parawixia bistriata spider venom. Mol Pharmacol 72: 1228-1237. doi: 10.1124/mol.107.037127
![]() |
[156] |
Fontana ACK, Guizzo R, Beleboni RD, et al. (2003) Purification of a neuroprotective component of Parawixia bistriata spider venom that enhances glutamate uptake. Br J Pharmacol 139: 1297-1309. doi: 10.1038/sj.bjp.0705352
![]() |
[157] |
Xing XC, Chang LC, Kong QM, et al. (2011) Structure-activity relationship study of pyridazine derivatives as glutamate transporter EAAT2 activators. Bioorg Med Chem Lett 21: 5774-5777. doi: 10.1016/j.bmcl.2011.08.009
![]() |
1. | Yunhui Zhang, Yongquan Zhou, Shuangxi Chen, Wenhong Xiao, Mingyu Wu, Bald eagle search algorithm for solving a three-dimensional path planning problem, 2024, 21, 1551-0018, 2856, 10.3934/mbe.2024127 | |
2. | João Carvalho, Tiago Pinto, Juan M. Home-Ortiz, Brigida Teixeira, Zita Vale, Ruben Romero, 2023, Chapter 25, 978-3-031-38317-5, 245, 10.1007/978-3-031-38318-2_25 |
Cluster | Problem | Features |
1 | One Max | Low mutation values and crossover does not have high values |
2 | One Max | All algorithms with mutation value of.5 |
3 | One Max | Mutation and crossover values are proportionally inverse |
1 | TSP | The mutation is low |
2 | TSP | The mutation has intermediate values |
3 | TSP | The mutation has the highest values |
1 | 3-SAT | The crossover has intermediate and low values, and mutation or too low or too high. |
2 | 3-SAT | Is not characterized for a special configuration |
3 | 3-SAT | Mutation has intermediate and high values |
Algorithm | Problem | Population Value | Cross Over Value | Mutation Value |
GA9 | One-Max | 300 | 0.9 | 0.1 |
GA18 | One-Max | 300 | 0.9 | 0.5 |
GA27 | One-Max | 300 | 0.9 | 0.9 |
GA1 | Dynamic One-Max | 50 | 0.1 | 0.1 |
GA10 | Dynamic One-Max | 50 | 0.1 | 0.5 |
GA19 | Dynamic One-Max | 50 | 0.1 | 0.9 |
GA19 | TSP | 50 | 0.1 | 0.9 |
GA20 | TSP | 150 | 0.1 | 0.9 |
GA21 | TSP | 300 | 0.1 | 0.9 |
GA11 | Dynamic TSP | 150 | 0.1 | 0.5 |
GA20 | Dynamic TSP | 150 | 0.1 | 0.9 |
GA21 | Dynamic TSP | 300 | 0.1 | 0.9 |
GA1 | 3-SAT | 50 | 0.1 | 0.1 |
GA10 | 3-SAT | 50 | 0.1 | 0.5 |
GA20 | 3-SAT | 150 | 0.1 | 0.9 |
Algorithm | Best | Average | Mode | Minimum |
GA1 | 50 | 47.5922 | 50 | 46.64 |
GA9 | 50 | 48.2436 | 50 | 47.66 |
GA10 | 50 | 47.7099 | 50 | 47.28 |
GA11 | 50 | 45.2242 | 49 | 40.6667 |
GA18 | 50 | 48.252 | 50 | 48.0067 |
GA19 | 50 | 47.7065 | 50 | 47.6082 |
GA21 | 50 | 30.046166 | 47 | 27.57 |
GA27 | 50 | 48.2448 | 50 | 48.0633 |
GA Weibull | 50 | 32.8611 | 50 | 27.05 |
Algorithm | Best | Average | Mode | Minimum |
GA1 | 49 | 45.2046 | 49 | 42.1 |
GA9 | 50 | 36.8534 | 50 | 24.2 |
GA10 | 49 | 45.3747 | 49 | 42.87 |
GA11 | 48 | 44.0205 | 43 | 37.5267 |
GA18 | 50 | 36.8391 | 50 | 24.7533 |
GA19 | 49 | 45.3738 | 49 | 42.83 |
GA21 | 50 | 36.1885 | 50 | 22.2767 |
GA27 | 50 | 36.8146 | 50 | 24.8033 |
GA Weibull | 50 | 49.4188 | 50 | 45.64 |
Algorithm | Best | Average | Mode |
GA1 | 76.8119 | 89.5805 | 76.8119 |
GA9 | 114.0008 | 146.4622 | 126.7976 |
GA10 | 76.1759 | 88.05 | 76.1759 |
GA11 | 76.3119 | 88.0503 | 76.3119 |
GA18 | 109.7748 | 146.1850 | 163.4686 |
GA19 | 75.5088 | 87.5741 | 75.5088 |
GA21 | 75.5088 | 87.6917 | 75.5088 |
GA27 | 107.2962 | 146.1011 | 125.7882 |
GA Weibull | 75.3097 | 80.4664 | 79.1822 |
Algorithm | Best | Average | Mode |
GA1 | 63.4885 | 97.7229 | 87.438 |
GA9 | 106.0332 | 173.1659 | 144.1324 |
GA10 | 66.1179 | 95.8859 | 78.0715 |
GA11 | 66.211 | 94.5666 | 76.374 |
GA18 | 94.6573 | 173.7578 | 214.5497 |
GA19 | 61.852 | 95.5419 | 70.1903 |
GA21 | 64.335 | 95.7753 | 77.8263 |
GA27 | 102.047 | 173.8176 | 140.1949 |
GA Weibull | 61.6235 | 75.8967 | 79.2923 |
Algorithm | Best | Average | Mode | Minimum |
GA1 | 91 | 89.9336 | 91 | 89.16 |
GA10 | 91 | 89.9899 | 91 | 89.5067 |
GA20 | 91 | 88.6487 | 91 | 88.1633 |
GA Weibull | 91 | 90.6304 | 90.84 | 88.44 |
Algorithm | Best | Average | Mode | Maximum |
GA1 | 179.67 | 177.9916 | 179 | 183 |
GA10 | 180.18 | 178.4567 | 180 | 183 |
GA20 | 180.19 | 176.6554 | 180 | 183 |
GA Weibull | 108.02 | 17.7718 | 180 | 184 |
Algorithm | Best | Average | Mode | Maximum |
GA1 | 53 | 52.4398 | 53 | 53 |
GA10 | 53 | 52.4388 | 53 | 53 |
GA20 | 53 | 51.5941 | 53 | 53 |
GA Weibull | 53 | 52.8658 | 53 | 53 |
Algorithm | Best | Average | Mode | Maximum |
GA1 | 105.14 | 103.9322 | 105 | 106 |
GA10 | 105.04 | 103.8655 | 105 | 106 |
GA20 | 105.13 | 105.5007 | 105 | 106 |
GA Weibull | 104.86 | 104.6524 | 105 | 106 |
Algorithm | Best | Average | Mode | Maximum |
GA1 | 75.99 | 75.1317 | 76 | 76 |
GA10 | 76 | 73.8836 | 76 | 76 |
GA20 | 76 | 75.0646 | 76 | 76 |
GA Weibull | 75.92 | 75.7808 | 76 | 76 |
Cluster | Problem | Features |
1 | One Max | Low mutation values and crossover does not have high values |
2 | One Max | All algorithms with mutation value of.5 |
3 | One Max | Mutation and crossover values are proportionally inverse |
1 | TSP | The mutation is low |
2 | TSP | The mutation has intermediate values |
3 | TSP | The mutation has the highest values |
1 | 3-SAT | The crossover has intermediate and low values, and mutation or too low or too high. |
2 | 3-SAT | Is not characterized for a special configuration |
3 | 3-SAT | Mutation has intermediate and high values |
Algorithm | Problem | Population Value | Cross Over Value | Mutation Value |
GA9 | One-Max | 300 | 0.9 | 0.1 |
GA18 | One-Max | 300 | 0.9 | 0.5 |
GA27 | One-Max | 300 | 0.9 | 0.9 |
GA1 | Dynamic One-Max | 50 | 0.1 | 0.1 |
GA10 | Dynamic One-Max | 50 | 0.1 | 0.5 |
GA19 | Dynamic One-Max | 50 | 0.1 | 0.9 |
GA19 | TSP | 50 | 0.1 | 0.9 |
GA20 | TSP | 150 | 0.1 | 0.9 |
GA21 | TSP | 300 | 0.1 | 0.9 |
GA11 | Dynamic TSP | 150 | 0.1 | 0.5 |
GA20 | Dynamic TSP | 150 | 0.1 | 0.9 |
GA21 | Dynamic TSP | 300 | 0.1 | 0.9 |
GA1 | 3-SAT | 50 | 0.1 | 0.1 |
GA10 | 3-SAT | 50 | 0.1 | 0.5 |
GA20 | 3-SAT | 150 | 0.1 | 0.9 |
Algorithm | Best | Average | Mode | Minimum |
GA1 | 50 | 47.5922 | 50 | 46.64 |
GA9 | 50 | 48.2436 | 50 | 47.66 |
GA10 | 50 | 47.7099 | 50 | 47.28 |
GA11 | 50 | 45.2242 | 49 | 40.6667 |
GA18 | 50 | 48.252 | 50 | 48.0067 |
GA19 | 50 | 47.7065 | 50 | 47.6082 |
GA21 | 50 | 30.046166 | 47 | 27.57 |
GA27 | 50 | 48.2448 | 50 | 48.0633 |
GA Weibull | 50 | 32.8611 | 50 | 27.05 |
Algorithm | Best | Average | Mode | Minimum |
GA1 | 49 | 45.2046 | 49 | 42.1 |
GA9 | 50 | 36.8534 | 50 | 24.2 |
GA10 | 49 | 45.3747 | 49 | 42.87 |
GA11 | 48 | 44.0205 | 43 | 37.5267 |
GA18 | 50 | 36.8391 | 50 | 24.7533 |
GA19 | 49 | 45.3738 | 49 | 42.83 |
GA21 | 50 | 36.1885 | 50 | 22.2767 |
GA27 | 50 | 36.8146 | 50 | 24.8033 |
GA Weibull | 50 | 49.4188 | 50 | 45.64 |
Algorithm | Best | Average | Mode |
GA1 | 76.8119 | 89.5805 | 76.8119 |
GA9 | 114.0008 | 146.4622 | 126.7976 |
GA10 | 76.1759 | 88.05 | 76.1759 |
GA11 | 76.3119 | 88.0503 | 76.3119 |
GA18 | 109.7748 | 146.1850 | 163.4686 |
GA19 | 75.5088 | 87.5741 | 75.5088 |
GA21 | 75.5088 | 87.6917 | 75.5088 |
GA27 | 107.2962 | 146.1011 | 125.7882 |
GA Weibull | 75.3097 | 80.4664 | 79.1822 |
Algorithm | Best | Average | Mode |
GA1 | 63.4885 | 97.7229 | 87.438 |
GA9 | 106.0332 | 173.1659 | 144.1324 |
GA10 | 66.1179 | 95.8859 | 78.0715 |
GA11 | 66.211 | 94.5666 | 76.374 |
GA18 | 94.6573 | 173.7578 | 214.5497 |
GA19 | 61.852 | 95.5419 | 70.1903 |
GA21 | 64.335 | 95.7753 | 77.8263 |
GA27 | 102.047 | 173.8176 | 140.1949 |
GA Weibull | 61.6235 | 75.8967 | 79.2923 |
Algorithm | Best | Average | Mode | Minimum |
GA1 | 91 | 89.9336 | 91 | 89.16 |
GA10 | 91 | 89.9899 | 91 | 89.5067 |
GA20 | 91 | 88.6487 | 91 | 88.1633 |
GA Weibull | 91 | 90.6304 | 90.84 | 88.44 |
Algorithm | Best | Average | Mode | Maximum |
GA1 | 179.67 | 177.9916 | 179 | 183 |
GA10 | 180.18 | 178.4567 | 180 | 183 |
GA20 | 180.19 | 176.6554 | 180 | 183 |
GA Weibull | 108.02 | 17.7718 | 180 | 184 |
Algorithm | Best | Average | Mode | Maximum |
GA1 | 53 | 52.4398 | 53 | 53 |
GA10 | 53 | 52.4388 | 53 | 53 |
GA20 | 53 | 51.5941 | 53 | 53 |
GA Weibull | 53 | 52.8658 | 53 | 53 |
Algorithm | Best | Average | Mode | Maximum |
GA1 | 105.14 | 103.9322 | 105 | 106 |
GA10 | 105.04 | 103.8655 | 105 | 106 |
GA20 | 105.13 | 105.5007 | 105 | 106 |
GA Weibull | 104.86 | 104.6524 | 105 | 106 |
Algorithm | Best | Average | Mode | Maximum |
GA1 | 75.99 | 75.1317 | 76 | 76 |
GA10 | 76 | 73.8836 | 76 | 76 |
GA20 | 76 | 75.0646 | 76 | 76 |
GA Weibull | 75.92 | 75.7808 | 76 | 76 |