International Journal of Engineering and Advanced Technology (IJEAT) ISSN: 2249 – 8958, Volume-9 Issue-2, December, 2019 445 Published By: Blue Eyes Intelligence Engineering & Sciences Publication Retrieval Number: B3136129219/2019©BEIESP DOI: 10.35940/ijeat.B3136.129219 Abstract: Sequence alignment is a significant facet in the bio-informatics research field for the molecular sequence analysis. Arrangement of two biological sequences by maximizing the similarities between the sequences by incorporating and adjusting gaps is Pairwise Sequence Alignment (PSA). Arrangement of multiple sequences is Multiple Sequence Alignment (MSA). Though Dynamic programming can produce optimal sequence alignment for PSA it suffers from a problem when multiple optimal paths are present and trace back is required. Back tracking becomes complex and it is also not suitable for MSA. So many meta-heuristic algorithms like Genetic Algorithm (GA) and Differential Evolutionary Algorithm (DE) are developed in the recent years to resolve the issue of optimization. Both GA and DE are used to produce optimal sequence alignment. But Compared to GA, DE is able to produce more optimal sequence alignment. To further enhance the performance of DE a new mutant is proposed by considering best, worst and a random candidate solution and applied on DE. It is named as New Differential Evolutionary Algorithm (NDE). By taking the test sequences from a bench mark data set “prefab4ref” tests are performed on GA, All DE mutants and NDE and it is observed that the proposed algorithm NDE outperformed all the other algorithms. Keywords: Sequence Alignment, Biological Sequences, Pairwise Sequence Alignment, Multiple Sequence Alignment, Genetic Algorithm, Differential Evolutionary Algorithm. I. INTRODUCTION Biological Informatics, in brief Bioinformatics is a combination of Biology, Computer Science and Information Technology. Now-a-days it is called as Computational Biology by many scientists. In order to solve so many biological problems, Scientists are continuously striving to design new algorithms [1]. Many bioinformatics tools and databases were designed and developed by scientists to analyse the biological data and to store the biological information. Bioinformatics covers so many areas like genetics, proteomics etc. One of the most prominent applications of the Bioinformatics is Sequence Analysis and Sequence Alignment. Sequence Alignment is mainly for the determination of analogous regions with in the specified Revised Manuscript Received on December 08, 2019 Lakshmi Naga Jayaprada. Gavarraju, Assoc.Prof, Dept. of Computer Science & Engineering, Narasaraopeta Engineering College [Autonomous], Narasaraopet, Guntur(Dt), A.P., India. Kanadam Karteeka Pavan, Professor & Head Department of Computer Applications, R.V.R.& J.C.College of Engineering [Autonomous], Chowdavaram , Guntur , A.P., India. biological sequences like nucleotide or protein sequences. Identifying the analogous areas within the specified biological sequences is for the purpose of finding functional similarity or structural similarity or to evolve evolutionary relationships among the specified sequences. Sequence alignment is majorly of two varieties based on number of sequences: Pair-wise and Multiple Sequence Alignments. Aligning two biological sequences is called PSA and Aligning multiple biological sequences is called MSA. Sequence Alignment is divided into two categories depending on the type of alignment: Global [2] and Local [3] Sequence Alignments. Performing sequence alignment on amino acid sequences is more appropriate than performing the sequence alignment on nucleotide sequences. It is because amino acid sequences (protein sequences) consist of functional and structural information [4]. Many scoring functions are utilized to find the similarity or identity among the sequences. When comparing the nucleotide sequences a simple scoring function called Identity Score (IS), can be used, where similar nucleotide bases is assigned a positive score, dissimilar a negative score and for a gap less negative score is assigned. Another Scoring function called Column Score (CS) can also be used, in which identical nucleotide bases are present in a single column a value of ‘1’ is assigned otherwise a value of ‘0’ is assigned. For protein sequences another scoring function called Similarity Score (SS) can also be used along with scoring functions like IS and CS. In SS, amino acids with analogous physiochemical properties are assigned a value based substitution matrices like Point Accepted Mutation (PAM) [5] and BLOcked Substitution Matrix (BLOSUM) [6]. A variety of PSA techniques are available to produce best alignment to two given sequences both local and global. To produce optimal alignment of the two given sequences, previously Dynamic programming was used. A Dynamic programming algorithm “Smith-Waterman algorithm” [7] is used for local sequence alignment and “Needleman-Wunsch algorithm” [8] is used for global sequence alignment. Both the algorithms suffer from a drawback specifically when two or many optimal paths are generated and trace back is needed. Back tracking becomes complex [9, 10]. So, many scientists tried to use nature inspired optimization algorithms. Genetic Algorithm (GA) is an optimization algorithm to solve the problem of sequence alignment to produce the optimal alignment. A multi objective GA was developed by Taneda for PSA of RNA sequence alignment [11]. Notredame et.al., developed an algorithms for the optimal alignment of RNA sequences by GA called RAGA and another algorithm Parallel GA (PRAGA) [12]. Cedric Pairwise Sequence Alignment by Differential Evolutionary Algorithm with New Mutation Strategy Lakshmi Naga Jayaprada.Gavarraju, K. Karteeka Pavan