Muhammad Akmal bin Remli

Automated biological pathway knowledge retrieval based on semantic web services composition and AI Planning

This paper presents the experience gained on semantic web service composition technique applied to the bioinformatics domain. Specifically, the approach presented here consists of knowledge retrieval perspective in biological pathway. Semantic web services, annotated with domain ontology are used to describe services for pathway knowledge retrieval for Kyoto Encyclopedia of Gene and Genomes (KEGG) database. Retrieving knowledge can be seen as high level goals and the tasks involved can be decomposed into subtask to achieve the specified goals. We execute the composition of service by treating composition as planning problem using Hierarchical Task Network (HTN) planning system based on Simple Hierarchical Order Planner 2 (SHOP2). The approach for plan (task) decomposition using SHOP2 is implemented in automated way. We investigate the effectiveness of this approach by applying real world scenario in pathway information retrieval for Lactococcus Lactis (L. lactis) organism where biologists need to find out the pathway description from the given specific gene of interest.

An enhanced scatter search with combined opposition-based learning for parameter estimation in large-scale kinetic models of biochemical systems

An enhanced scatter search (eSS) with combined opposition-based learning algorithm is proposed to solve large-scale parameter estimation in kinetic models of biochemical systems. The proposed algorithm is an extension of eSS with three important improvements in terms of: reference set (RefSet) formation, RefSet combination, and RefSet intensification. Due to the difficulty in estimating kinetic parameter values in the presence of noise and large number of parameters (high-dimension), the aforementioned eSS mechanisms have been improved using combination of quasi-opposition and quasi-reflection, which were under the family of opposition-based learning scheme. The proposed algorithm is tested using one set of benchmark function each from large-scale global optimization (LSGO) problem as well as parameter estimation problem. The LSGO problem consists of 11 functions with 1000 dimensions. For parameter estimation, around 116 kinetic parameters in Chinese hamster ovary (CHO) cells and central carbon metabolism of E. coli are estimated. The results revealed that the proposed algorithm is superior to eSS and other competitive algorithms in terms of its efficiency in minimizing objective function value and having faster convergence rate. The proposed algorithm also required lower computational resources, especially number of function evaluations performed and computation time. In addition, the estimated kinetic parameter values obtained from the proposed algorithm produced the best fit to a set of experimental data.

An improved hybrid of particle swarm optimization and the gravitational search algorithm to produce a kinetic parameter estimation of aspartate biochemical pathways

Mathematical modelling is fundamental to understand the dynamic behavior and regulation of the biochemical metabolisms and pathways that are found in biological systems. Pathways are used to describe complex processes that involve many parameters. It is important to have an accurate and complete set of parameters that describe the characteristics of a given model. However, measuring these parameters is typically difficult and even impossible in some cases. Furthermore, the experimental data are often incomplete and also suffer from experimental noise. These shortcomings make it challenging to identify the best-fit parameters that can represent the actual biological processes involved in biological systems. Computational approaches are required to estimate these parameters. The estimation is converted into multimodal optimization problems that require a global optimization algorithm that can avoid local solutions. These local solutions can lead to a bad fit when calibrating with a model. Although the model itself can potentially match a set of experimental data, a high-performance estimation algorithm is required to improve the quality of the solutions. This paper describes an improved hybrid of particle swarm optimization and the gravitational search algorithm (IPSOGSA) to improve the efficiency of a global optimum (the best set of kinetic parameter values) search. The findings suggest that the proposed algorithm is capable of narrowing down the search space by exploiting the feasible solution areas. Hence, the proposed algorithm is able to achieve a near-optimal set of parameters at a fast convergence speed. The proposed algorithm was tested and evaluated based on two aspartate pathways that were obtained from the BioModels Database. The results show that the proposed algorithm outperformed other standard optimization algorithms in terms of accuracy and near-optimal kinetic parameter estimation. Nevertheless, the proposed algorithm is only expected to work well in small scale systems. In addition, the results of this study can be used to estimate kinetic parameter values in the stage of model selection for different experimental conditions.

Metaheuristic Optimization for Parameter Estimation in Kinetic Models of Biological Systems - Recent Development and Future Direction

Background: Kinetic models with predictive ability are important to be used in industrial biotechnology. However, the most challenging task in kinetic modeling is parameter estimation, which can be addressed using metaheuristic optimization methods. The methods are utilized to minimize scalar distance between model output and experimental data. Due to highly nonlinear nature of biological systems and large number of kinetic parameters, parameter estimation becomes difficult and time consuming. Methods: This paper provides a review on recent development of parameter estimation methods, which has received increasing attention in the field of systems biology. The development of metaheuristic optimization methods is mostly focused in this review along with the development of large-scale kinetic models. Results: Although a plethora of methods have been applied to the problem of parameter estimation, recent results show that most of the successful approaches are those based on hybrid methods and parallel strategies. In addition, the current software used for parameter estimation and the sources of biological data for kinetic modeling are also described in this review. This review also presents future direction in parameter estimation to meet current industrial demands, especially in systems biology applications. Conclusion: The development of numerous optimization methods for parameter estimation in kinetic models has brought much advancement in the application of systems biology. Currently, it seems that there are highly demanded for further development of efficient optimization methods to address the expansion of systems biology applications.

Cooperative enhanced scatter search with opposition-based learning schemes for parameter estimation in high dimensional kinetic models of biological systems

Abstract Industrial bioprocesses development nowadays is concerned with producing chemicals using yeast, bacteria and therapeutic proteins in mammalian cells. This involves the utilization of microorganism cells as factories and re-engineering them in silico. The tools that could facilitate this process are known as the kinetic models. Kinetic models of cellular metabolism are important in assisting researchers to understand the rational design of biological systems, predicting metabolites production, and improving bio-products development. However, the most challenging task in model development is parameter estimation, which is the process of identifying an unknown value of model parameters which provides the best fit between the model output and a set of experimental data. Due to the increased complexity and high dimensionality of the models, which are extremely nonlinear and contain large numbers of kinetic parameters, parameter estimation is known to be difficult and time-consuming. This study proposes a cooperative enhanced scatter search with opposition-based learning schemes (CeSSOL) for parameter estimation in large-scale biology models. The method was executed in parallel with the proposed cooperative mechanism in order to exchange information (kinetic parameters) between individual threads. Each thread consists of different parameters settings that enhance the systemic properties in obtaining the global minimum. The performance of the proposed method was assessed against two large-scale microorganisms models using mammalian and bacteria cells. The results revealed that the proposed method recorded faster computation time compared to other methods. The study has also demonstrated that the proposed method can be used to provide more accurate and faster estimation of kinetic models, indicating the potential benefits of utilizing this method for expert systems of industrial biotechnology.

Parameter Estimation of Essential Amino Acids in Arabidopsis thaliana Using Hybrid of Bees Algorithm and Harmony Search

Mathematical models of metabolic processes are the cornerstone of computational systems biology. In model building, the task of parameter estimation is difficult due to the huge numbers of kinetics parameters involved. The common way of estimating the parameters is to formulate it as an optimization problem. Global optimization methods can be applied by minimizing the distance between experimental data and predicted models. This paper proposes the Hybrid of Bees Algorithm and Harmony Search (BAHS) to estimate the kinetics parameters of essential amino acid production in the aspartate metabolism for Arabidopsis thaliana. The performance of the BAHS is evaluated and compared with other algorithms. The results show that BAHS performed better as it improved the performance of the original BA by 60%. Meanwhile, it takes less computational time to estimate the kinetics parameters of essential amino acid production for Arabidopsis thaliana.

A Hybrid of Simple Constrained Artificial Bee Colony Algorithm and Flux Balance Analysis for Enhancing Lactate and Succinate in Escherichia Coli.

In the past decades, metabolic engineering has received great attention from different sectors of science due to its important role in enhancing the over expression of the target phenotype by manipulating the metabolic pathway. The advent of metabolic engineering has further laid the foundation for computational biology, leading to the development of computational approaches for suggesting genetic manipulation. Previously, conventional methods have been used to enhance the production of lactate and succinate in E. coli. However, these products are always far below their theoretical maxima. In this research, a hybrid algorithm is developed to seek optimal solutions in order to increase the overproduction of lactate and succinate by gene knockout in E. coli. The hybrid algorithm employed the Simple Constrained Artificial Bee Colony (SCABC) algorithm, using swarm intelligence as an optimization algorithm to optimize the objective function, where lactate and succinate productions are maximized by simulating gene knockout in E. coli. In addition, Flux Balance Analysis (FBA) is used as a fitness function in the SCABC algorithm to assess the growth rate of E. coli and the productivity of lactate and succinate. As a result of the research, the gene knockout list which induced the highest production of lactate and succinate is obtained.

A hybrid of Cuckoo Search and Minimization of Metabolic Adjustment to optimize metabolites production in genome-scale models

Metabolic engineering involves the modification and alteration of metabolic pathways to improve the production of desired substance. The modification can be made using in silico gene knockout simulation that is able to predict and analyse the disrupted genes which may enhance the metabolites production. Global optimization algorithms have been widely used for identifying gene knockout strategies. However, their productions were less than theoretical maximum and the algorithms are easily trapped into local optima. These algorithms also require a very large computation time to obtain acceptable results. This is due to the complexity of the metabolic models which are high dimensional and contain thousands of reactions. In this paper, a hybrid algorithm of Cuckoo Search and Minimization of Metabolic Adjustment is proposed to overcome the aforementioned problems. The hybrid algorithm searches for the near-optimal set of gene knockouts that leads to the overproduction of metabolites. Computational experiments on two sets of genome-scale metabolic models demonstrate that the proposed algorithm is better than the previous works in terms of growth rate, Biomass Product Couple Yield, and computation time.

Classification of colorectal cancer using clustering and feature selection approaches

Accurate cancer classification and responses to treatment are important in clinical cancer research since cancer acts as a family of gene-based diseases. Microarray technology has widely developed to measure gene expression level changes under normal and experimental conditions. Normally, gene expression data are high dimensional and characterized by small sample sizes. Thus, feature selection is needed to find the smallest number of informative genes and improve the classification accuracy and the biological interpretability results. Due to some feature selection methods neglect the interactions among genes, thus, clustering is used to group the similar genes together. Besides, the quality of the selected data can determine the effectiveness of the classifiers. This research proposed clustering and feature selection approaches to classify the gene expression data of colorectal cancer. Subsequently, a feature selection approach based on centroid clustering provide higher classification accuracy compared with other approaches.

K-means clustering with infinite feature selection for classification tasks in gene expression data

In the bioinformatics and clinical research areas, microarray technology has been widely used to distinguish a cancer dataset between normal and tumour samples. However, the high dimensionality of gene expression data affects the classification accuracy of an experiment. Thus, feature selection is needed to select informative genes and remove non-informative genes. Some of the feature selection methods, yet, ignore the interaction between genes. Therefore, the similar genes are clustered together and dissimilar genes are clustered in other groups. Hence, to provide a higher classification accuracy, this research proposed k-means clustering and infinite feature selection for identifying informative genes in the selected subset. This research has been applied to colorectal cancer and small round blue cell tumors datasets. Eventually, this research successfully obtained higher classification accuracy than the previous work.