Deep Learning Methods and Optimization Algorithms in Bioinformatics Field: A Review

Abstract

Bioinformatics is a multidisciplinary field which combines computer science, mathematics, statistics and also engineering to analyze, simulate and manipulate the biological data. However, researchers in bioinformatics field often suffer from the difficulties to encode the knowledge contain in biological data. It is not easy to analyze and manipulate biological data due to its hard nature. Furthermore, experimental in wet laboratory setting always required highly cost and took longer time to get the results. In this fast growing era, various computational methods have been developed in order to solve these problems as well as give better finding results with their fast performance. Deep learning is one of the computational methods that has been in spotlight and widely used in bioinformatics field due to its capability in predicting protein properties, gene ontology annotation and many more. Other than deep learning, optimization algorithms also can be considered as mostly used algorithms nowadays in bioinformatics field to assists deep learning in getting the best output from biological data. This paper will review several deep learning methods as well as optimization algorithms and their involvement in bioinformatics field. Summarization of advantages and drawbacks for both deep learning methods and optimization algorithms also will be visualized in the table forms at the end of every section.