Novel methods to optimize gene and statistic test for evaluation – an application for Escherichia coli

BMC Bioinformatics

Methodology Article

Ngo Quoc Viet¹  Tran Tuan-Anh²  Pham The Bao²  Le Thi Ly³ 

[1] Faculty of Information Technology, Ho Chi Minh City University of Pedagogy, 280 An Duong Vuong Street, Ward 4, District 5, Ho Chi Minh City, Vietnam;Faculty of Mathematics and Computer Science, VNUHCM-University of Science, 227 Nguyen Van Cu Street, District 5, Ho Chi Minh City, Vietnam;School of Biotechnology, VNUHCM-International University, Quarter 6, Linh Trung Ward, Thu Duc District, Ho Chi Minh City, Vietnam;

关键词: Gene optimization;    Neural network;    Bayes’ theorem;    Euclidean distance;    Codon usage bias;    Highly expressed gene;   

DOI  :  10.1186/s12859-017-1517-z

 received in 2016-04-23, accepted in 2017-02-01,  发布年份 2017

来源: Springer

【 摘 要 】

BackgroundSince the recombinant protein was discovered, it has become more popular in many aspects of life science. The value of global pharmaceutical market was $87 billion in 2008 and the sales for industrial enzyme exceeded $4 billion in 2012. This is strong evidence showing the great potential of recombinant protein. However, native genes introduced into a host can cause incompatibility of codon usage bias, GC content, repeat region, Shine-Dalgarno sequence with host’s expression system, so the yields can fall down significantly. Hence, we propose novel methods for gene optimization based on neural network, Bayesian theory, and Euclidian distance.ResultThe correlation coefficients of our neural network are 0.86, 0.73, and 0.90 in training, validation, and testing process. In addition, genes optimized by our methods seem to associate with highly expressed genes and give reasonable codon adaptation index values. Furthermore, genes optimized by the proposed methods are highly matched with the previous experimental data.ConclusionThe proposed methods have high potential for gene optimization and further researches in gene expression. We built a demonstrative program using Matlab R2014a under Mac OS X. The program was published in both standalone executable program and Matlab function files. The developed program can be accessed from http://www.math.hcmus.edu.vn/~ptbao/paper_soft/GeneOptProg/.

【 授权许可】

CC BY   
© The Author(s). 2017

【 图 表 】

12864_2017_3605_Article_IEq3.gif

12864_2017_3492_Article_IEq1.gif

12864_2015_2055_Article_IEq34.gif

12864_2017_4133_Article_IEq18.gif

12864_2017_3487_Article_IEq38.gif

12864_2017_4132_Article_IEq30.gif

12864_2017_4132_Article_IEq28.gif

12864_2017_4132_Article_IEq27.gif

12864_2016_2789_Article_IEq20.gif

12864_2017_3610_Article_IEq3.gif

12864_2017_4179_Article_IEq34.gif

12864_2017_3670_Article_IEq11.gif

12864_2017_4179_Article_IEq31.gif