Development And Assembly For Several DP Based Sequence Alignment Algorithm Components

Sequence alignment is a technique for obtaining functional,structural or evolutionary relationships between sequences by aligning genomic sequences and identifying sequence similarity regions.With the implementation of the Human Genome Project,the development of sequencing technology has produced a large number of original sequence data on biomolecules.How to efficiently process and analyze these data in the face of such rich genomic sequence data,and compare similar regions or the conservation site between two sequences,seeking sequence homology structure,revealing the problems of biological inheritance,variation and evolution,has become one of the main engine of sequence alignment algorithm research.At present,the research of comparison algorithms mostly focuses on specific problems in the domain of sequence similarity analysis or specific algorithm optimization,but less on the whole problem domain,it is difficult to obtain a higherlevel of abstraction and suitable for the whole sequence similarity analysis.The algorithm component library of the domain leads to some problems such as the redundancy of the sequence comparison algorithm and the error that the artificial selection algorithm may cause.It also makes it difficult for people to effectively understand the algorithm structure,and cannot guarantee the correct use of the algorithm,so much as the accuracy of the sequence similarity analysis results is reduced.Due to the specificity and low abstraction of existing algorithms,researchers not only need to spend a lot of time to learn and use such algorithms,which reduces the maintainability and reusability of the algorithms,and it is difficult to locate and solve the errors generated by the algorithms.The burden of sequence similarity analysis is aggravated.Through in-depth analysis of the domain of Dynamic Programming-based Pairwise Sequence Alignment Algorithm(DPPSAA),the DPPSAA domain feature model is designed and built by using the generative programming method,which mainly includes the feature model and feature interaction model.Then,the features in the above domain feature model were formally implemented as components by using the high abstract programming language Apla of PAR platform,and the implementation process was described in detail.A DPPSAA component library with high abstract based on Apla program was established,and the specific assembly steps of Needleman-Wunsch algorithm based on this component library were demonstrate.At the same time,using the C++ program conversion system of PAR platform,the above component library is converted into the corresponding C++ program code by manual and automatic combination,and the Needleman-Wunsch algorithm is assembled and implemented.The results show that the DPPSAA component library is practical and has certain advantages in algorithm learning and implementation,which makes it possible to formal assembly sequence alignment algorithms.Then,based on the DPPSAA component library,some component library components are improved and added to support the assembly of multiple sequence alignment domain algorithms based on DPPSAA component library,and the expanded component library is used to assemble and implement the star alignment algorithm.The experimental results show that the star alignment algorithm generated by the assembly has certain practicability,which not only verifies the reusability of the DPPSAA components library,but also expands the scope of application of the DPPSAA domain components.