A particle swarm optimization solution for challenging planted(l, d)-Motif problem

Abstract

In Bioinformatics, Planted (l, d)-Motif finding is an important and challenging problem, which has many applications. Generally, it is to locate recurring patterns in the promoter regions of co-expressed or co-regulated genes. As we can't expect the pattern to be exact matching copies owing to biological mutations, the motif finding turns to be an NPcomplete problem. By approximating the same in different aspects, scientists have provided many solutions in the literature. These solutions are either "exact" or "approximate". All the proposed exact solutions take exponential-time; they need more time to search for larger parameters l and d. The problems of Bioinformatics seldom need the exact optimum solution; rather what they need is robust, fast and near optimal solutions. Therefore, it is impractical to use an exact algorithm to search for large parameters of motifs in real biological dataset. In this paper, we have adopted the features of the Particle Swarm Optimization (PSO) with k-nearest neighbor algorithm to solve the Planted (l, d)-Motif Finding Problem. PSO is a global approximation optimization technique and has wide applications.It finds the global best solution by simply adjusting the trajectory of each individual towards its own best location and towards the best particle of the swarm at each generation. We have performed some experiments on synthetic data by increasing number of sequences and the length of the sequences for different (l, d)-Motifs for the following data sets: general instances (15) and finally, we have applied our proposed method for real biological sequences. From the experimental results we observe that the proposed algorithm is more efficient and accurate compared to existing approximation algorithms and even it works better for larger motif instances.

Quorum Planted Motif Discovery and Motif Finding Using S2f and Fff Algorithms

“…The performance of the proposed work on other existing algorithms is compared by using the CRP-dataset as a reference and is shown in Table 4. Here the proposed algorithm is compared with the algorithms Projection [12], MEME [9], PSOKNN [25], and PSORPS [26]. The motif sequence logo [27] is represented in Fig.…”

Discovery of DNA Motif Utilising an Integrated Strategy Based on Random Projection and Particle Swarm Optimization

“…where the elements in (14) are defined as follows: In the dataset, the length of each sequence is 105, and the length of transcription factor binding sites is 22. To compare the performance of the proposed algorithm, we select motif consensus-based random projection algorithm (Projection) [19], position weight matrix-based statistical algorithm (MEME) [17], and particle swarm optimizationbased intelligent algorithm (PSOKNN) [29] for comparisons. Table 1 presents the comparison results.…”

“…Recently, population-based swarm intelligence algorithms have been proposed to address PMP problem [23][24][25][26]. Representative methods of this kind include PSOGAP [27] and PSO-MAC [28], and PSO-KNN [29]. The PSOGAP enumerates all the l-mers and constructs a similarity matrix for the l-mers.…”