Xylitol production of E. coli using deep neural network and firefly algorithm

Abstract

The emergence of deep learning as a technique forms a part of artificial intelligence give a huge contribution in machine learning towards the development of powerful tools. Deep learning is potentially being well suited in genomics representations which enable distributed representations’ data from multiple processing layers. Practically, deep learning is capable in demonstrating abstraction within the cell in genomic analysis with high predictive power reinforces leads this research. The enhancement of deep neural network in representing genome-scale data into mathematical model allows predictive analysis to be conducted. This work aims to investigate biological process within E. coli to explore genomics representation in identifying target microbial production. Furthermore, the use of firefly algorithm prevents it from getting stuck at local optima in finding optimal solution during network training. The outcome of this study contributes in identifying the effects of genetic perturbation towards xylitol production of selective metabolic pathway in metabolic network of E. coli.