Document Type : Research Paper

Authors

• Lala Septem Riza ¹
• Muhammad Ilham Nurfathiya ¹
• Jajang Kusnendar ¹
• Khyrina Airin Fariza Abu Samah ²

¹ Department of Computer Science Education, Universitas Pendidikan Indonesia, Indonesia

² Faculty of Computer and Mathematical Sciences, University Teknologi MARA Cawangan Melaka Kampus Jasin, Melaka, Malaysia

Abstract

The objective of this research is to design and implement a computational model to determine DNA barcodes by utilizing the Particle Swarm Optimization (PSO) algorithms implemented on Big Data Platforms, namely Apache Hadoop and Apache Spark. The steps are as follows: (i) inputting DNA sequences to Hadoop Distributed File System (HDFS) in Apache Hadoop, (ii) pre-processing data, (iii) implementing PSO by utilizing the User Defined Function (UDF) in Apache Spark, (iv) collecting results and saving to HDFS. After obtaining the computational model, two following simulations have been done: the first scenario is using 4 cores and several worker nodes, meanwhile, the second one consists of a cluster with 2 worker nodes and several cores. In terms of computational time, the results show a significant acceleration between standalone and big data platforms with both experimental scenarios. This study proves that the computational model built on the big data platform shows the development of features and acceleration of previous research.

Keywords

• Big data
• Algorithm
• Particle swarm optimization
• Similarity check
• Motif discovery
• DNA barcoding