A parallel cooperative team of multiobjective evolutionary algorithms for motif discovery

The DNA motif discovery (MD) problem is the main challenge of genome biology, and its importance is directly proportional to increasing sequencing technologies. MD plays a vital role in the identification of transcription factor binding sites that help in learning the mechanisms for regulation of gene expression. Metaheuristic algorithms are promising techniques for eliciting motif from DNA genomic sequences, but often fail to demonstrate robust performance by overcoming the inherent challenges in complex gene sequences, making search environment extremely non-convex for optimization methods. This paper proposes a novel modified Henry gas solubility optimization (MHGSO) algorithm for motif discovery which elicits a functional motif in DNA genomic sequences. In our approach, a new stage that captures the main characteristics of the motifs in DNA sequences is proposed, and MHGSO imitates the motifs characteristics for accurate detection of target motif. The performance of the MHGSO algorithm is validated using both synthetic and real datasets. Results confirm the stability and superiority of the proposed algorithm compared to state-of-the-art algorithms including MEME, DREME, XXmotif, PMbPSO, and MACS. Based on several evaluation matrices, MHGSO outperforms the competitor techniques in terms of nucleotide-level correlation coefficient, recall, precision, F-score, Cohen’s Kappa, and statistical validation measures.

We propose a new algorithm, PMS6, for the (l,d)-motif discovery problem in which we are to find all strings of length l that appear in every string of a given set of strings with at most d mismatches. The run time ratio PMS5/PMS6, where PMS5 is the fastest previously known algorithm for motif discovery in large instances, ranges from a high of 2.20 for the (21,8) challenge instances to a low of 1.69 for the (17,6) challenge instances. Both PMS5 and PMS6 require some amount of pre-processing. The pre-processing time for PMS6 is 34 times faster than that for PMS5 for (23,9) instances. When pre-processing time is factored in, the run time ratio PMS5/PMS6 is as high as 2.75 for (13,4) instances and as low as 1.95 for (17,6) instances.

Motif discovery problem is crucial for understanding the structure and function of gene expression. Over the past decades, many attempts using consensus and probability training model for motif finding are successful. However, the most existing motif discovery algorithms are still time-consuming or easily trapped in a local optimum. To overcome these shortcomings, in this paper, we propose an entropy-based position projection algorithm, called EPP, which designs a projection process to divide the dataset and explores the best local optimal solution. The experimental results on real DNA sequences, Tompa data, and ChIP-seq data show that EPP is advantageous in dealing with the motif discovery problem and outperforms current widely used algorithms.

A cis-regulatory element is a non-coding DNA sequence in or near a gene required for its expression, often containing transcription factors binding sites (TFBS). Identifying new TFBSs is fundamental to understand the regulation process of these genes. The optimization problem responsible for these findings is the Motif Discovery Problem (MDP), which is formulated as a short DNA pattern discovery problem. In addition, the MDP is an NP-hard optimization problem since it has to find short strings mixed among a large amount of biological information. In this paper, we propose and study a new hybrid multiobjective evolutionary algorithm based on Differential Evolution (DE) named Hybrid Differential Evolution with Pareto Tournaments (H-DEPT). To demonstrate the superiority of our proposed algorithm on a biological level, we compare the predictions made with those predicted by MEME, BioProspector and BioOptimizer when solving a set of four real sequence datasets. Finally, by using OpenMP, we also demonstrate that this algorithm is highly suitable to be parallelized. As we will see, we have performed several experiments with multi-core systems of 2, 4, and 8 cores; obtaining good parallel results.

We present a Bayesian motif discovery (BMD) algorithm for detecting an unknown number of instances of a motif in a given set of sequences. The algorithm models a motif with a position weight matrix (PWM), which is estimated along with the motif discovery process. This technique is flexible enough to enable other discovery algorithms' results to be used as input. The method is based on a sequential Monte Carlo algorithm, where the state to be estimated consists of the number of instances in each sequence and their initial positions. The accuracy of the proposed method is compared with other profile-based discovery algorithms. BMD is shown to perform statistically better than MEME and BioProspector in applications ranging from synthetic data to genomic motif finding of Din serine recombinases. In the case of site-specific recombinase target discovery, BMD-inferred motif is found to be the only functionally accurate from the underlying biochemical mechanism standpoint.

Multiobjective optimization is successfully applied in many biological problems. Currently, most biological problems require to optimize more than one single objective at the same time, resulting in Multiobjective Optimization Problems (MOP). In the last years, multiple metaheuristics have been successfully used to solve optimization problems. However, many of them are designed to solve problems with only one objective function. In this work, we study several multiobjective adaptations to solve one of the most important biological problems, the Motif Discovery Problem (MDP). MDP aims to discover novel Transcription Factor Binding Sites (TFBS) in DNA sequences, maximizing three conflicting objectives: motif length, support, and similarity. For this purpose, we have used the Artificial Bee Colony algorithm, a novel Swarm Intelligence algorithm based on the intelligent behavior of honey bees. As we will see, the use of one or another multiobjective adaptation causes significant differences in the results.

Motif finding problems, abstracted as the planted (l, d)-motif finding problem, are a major task in molecular biology--finding functioning units and genes. In 2002, the random projection algorithm was introduced to solve the challenging (15, 4)-motif finding problem by using randomly chosen templates. Two years later, a so-called uniform projection algorithm was developed to improve the random projection algorithm by means of low-dispersion sequences generated by coverings. In this article, we introduce an improved projection algorithm called the low-dispersion projection algorithm, which uses low-dispersion sequences generated by developed almost difference families. Compared with the random projection algorithm, the low-dispersion projection algorithm can solve the (l, d)-motif finding problem with fewer templates without decreasing the success rate.

With the fast digitalization of our society, mining patterns from large time series data is increasingly becoming a critical problem for a wide range of big data applications. Motif and discord discovery algorithms, which offer effective solutions to identify repeatedly appearing and abnormal patterns, respectively, are fundamental building blocks for time series processing. Both approaches, however, can be time extremely consuming when handling large time series due to the subsequence-based computations of distance similarity metrics. In this article, we show that the highly involved subsequence-based computations can actually be decomposed into a few fine-grained computing patterns for efficient data parallel computing. By developing highly efficient GPU algorithms for such basic patterns and effectively composing such patterns, we are able to solve both motif and discord discovery problems under euclidean and DTW distance metrics in a unified GPU acceleration framework. Extensive experiments prove that the proposed framework outperforms pruned CPU algorithms by up to three orders of magnitude. Our work paves the foundation of building GPU acceleration frameworks for large-scale time series datasets.

Solving optimization problems with multiple (often conflicting) objectives is, generally, a very difficult goal. Evolutionary algorithms (EAs) were initially extended and applied during the mid-eighties in an attempt to stochastically solve problems of this generic class. During the past decade, a variety, of multiobjective EA (MOEA) techniques have been proposed and applied to many scientific and engineering applications. Our discussion's intent is to rigorously define multiobjective optimization problems and certain related concepts, present an MOEA classification scheme, and evaluate the variety of contemporary MOEAs. Current MOEA theoretical developments are evaluated; specific topics addressed include fitness functions, Pareto ranking, niching, fitness sharing, mating restriction, and secondary populations. Since the development and application of MOEAs is a dynamic and rapidly growing activity, we focus on key analytical insights based upon critical MOEA evaluation of current research and applications. Recommended MOEA designs are presented, along with conclusions and recommendations for future work.

The motif finding problem is one of the important and challenging problems in bioinformatics. A variety of sequential algorithms have been proposed to find exact motifs, but the running time is still not suitable due to high computational complexity of finding motifs. In this paper we parallelize three efficient sequential algorithms which are HEPPMSprune, PMS5 and PMS6. We implement the algorithms on a Dual Quad-Core machine using openMP to measure the performance of each algorithm. Our experiment on simulated data show that: (1) the parallel PMS6 is faster than the other algorithms in case of challenging instances, while the parallel HEPPMSprune is faster than the other algorithms in most of solvable instances; (2) the scalability of parallel HEPPMSprune is linear for all instances, while the scalability of parallel PMS5 and PMS6 is linear in case of challenging instances only; (3) the memory used by HEPPMSprune is less than that of the other algorithms.

In this work we propose the application of a Swarm Intelligence (SI) algorithm to solve the Motif Discovery Problem (MDP), applied to the specific task of discovering novel Transcription Factor Binding Sites (TFBS) in DNA sequences. In the last years there have appeared many new evolutionary algorithms based on the collective intelligence. Finding TFBS is crucial for understanding the gene regulatory relationship but, motifs are weakly conserved, and motif discovery is an NP-hard problem. Therefore, the use of such algorithms can be a good way to obtain quality results. The chosen algorithm is the Artificial Bee Colony (ABC), it is an optimization algorithm based on the intelligent foraging behaviour of honey bee swarm. To solve the MDP we have applied multiobjective optimization and consequently, we have adapted the ABC to multiobjective problems, defining the Multiobjective Artificial Bee Colony (MOABC) algorithm. New results have been obtained, that significantly improve those published in previous researches.

Many real-world optimization problems are characterized by multiple conflicting objectives, which are known as multi-objective optimization problems (MOPs). In the last two decades, evolutionary algorithms have shown promising performance in solving various MOPs, and a large number of multi-objective evolutionary algorithms (MOEAs) have been proposed. In order to determine the most suitable MOEA for a specific MOP, it is usually necessary to perform experiments to compare the performance of multiple candidate MOEAs. In 2017, an evolutionary multi-objective optimization platform was proposed by us, called PlatEMO, which provides the source codes of many state-of-the-art MOEAs and helps researchers perform batch experiments on these MOEAs. In this work, we illustrate the method of using the newest version of PlatEMO to solve MOPs in applications, by means of a case study on the feature selection problem, which is an important and difficult task in machine learning and data mining. This paper details the method of adding the feature selection problem to PlatEMO, and presents the experimental results of eight MOEAs on nine datasets in feature selection.

Challenging multi-modal optimization problems have been very successfully solved by evolutionary computation (EC) techniques. To date, many methods have been proposed on evolutionary optimization for both single and multiobjective large scale problems. In the age of Big Data, there is an urge to take evolutionary optimization techniques to the next level for solving problems with even larger scales: thousands and millions of variables. These problems arise in many domains ranging from bioinformatics, to neuroscience and social simulations. In this paper, we investigate the use of EC to solve Big electroencephalography (EEG) data optimization problems with thousands of variables. The optimization problem attempts to identify maximum information that should be kept from a signal while minimizing the artifact. The high level of epistasis inherent in a signal can slow down the evolution. Therefore, we investigate the advantages of optimizing the problem in the frequency domain with different thresholds as opposed to the time domain. We propose synthetic EEG data sets of various scale and noise level. These data sets were the basis for the Optimization of Big Data 2015 Competition (BigOpt), CEC 2015. Two state-of-art multiobjective evolutionary algorithms (MOEAs) were evaluated. The results of this work suggest that frequency representation of the signals facilitates dimensionality reduction for big scale optimization of time series data, and hence provides faster and better quality solutions for EEG data cleaning. Moreover, the results suggest that existing state-of-art multiobjective evolutionary computation methods are extremely slow. Methods that can optimize the problem faster and with high quality are needed.

Engineering design problems are often characterized by multiple, conflicting objectives. Multiobjective evolutionary algorithms (MOEAs) can optimize these complex problems using embedded simulation models to calculate objective function values without simplifying assumptions, which are often required by traditional search algorithms. This study contributes a demonstration of how the performance of the MOEA can be improved through iterative reformulations of the problem objectives and constraints. We illustrate this iterative design approach using a case study of engineered injection and extraction (EIE), a strategy developed to enhance contaminant degradation during in situ groundwater remediation. During EIE, clean water is injected or extracted at wells surrounding a contaminated groundwater plume following the one-time injection of a treatment chemical. This sequence of injections and extractions reconfigures the plume and increases its contact with the treatment chemical, which ultimately increases the rate of contaminant degradation. The MOEA is used to optimize the EIE design, which consists of the sequence of pumping rates and well locations that dictate the plume reconfiguration. The optimization problem is a challenging one, characterized by uncertainty from randomness associated with hydrodynamic dispersion of the plume. The objectives and constraints of the remediation are refined iteratively during the optimization process, after considering tradeoffs between objectives for design solutions generated by the MOEA. Tradeoff data not only informs the design of solutions, it provides valuable information about the structure and conflicts of the design problem, which represents crucial support for stakeholders in their decision-making analyses for in situ groundwater remediation project planning with EIE.

Distribution system (DS) service restoration is a very complex multi-objective and multi-constraint optimization problem, which requires high-quality Pareto fronts for helping the DS operators' work. This paper proposes a multi-objective evolutionary algorithm (MOEA) that combines the advantages of MOEAs in Tables with the properties of the Non-dominated Sorting Genetic Algorithm-II (NSGA-II) to recover Pareto fronts of the service problem. Its main differentials are its ability of recovering high-quality Pareto fronts, even in large-scale DSs, and of prioritizing switching operations in Remotely Controlled Switches (RCSs), independently of amount of RCSs. In opposite of Manually Controlled Switches (MCSs), RCS can be faster and cheapest operated from operating center. Several tests were carried out to validate and to compare the proposed MOEA to three MOEAs published in literature. Four metrics were used for assessing the quality of Pareto fronts generated and a Welch's t hypothesis test enabled the comparison of MOEAs' performance. Test results indicate the proposed MOEA outperforms those three MOEAs published in literature.