Metabolic pathway synthesis based on predicting compound transformable pairs by using neural classifiers with imbalanced data handling
作者：
Highlights：
• Neural networks for predicting transform-ability of E.coli metabolite pairs.
• Average and projected features were two proposed schemes for each metabolite pair.
• Handling imbalanced training data by balancing boundary data of different classes.
• Our outperforming results in terms of accuracy and recovering pathway.
摘要
•Neural networks for predicting transform-ability of E.coli metabolite pairs.•Average and projected features were two proposed schemes for each metabolite pair.•Handling imbalanced training data by balancing boundary data of different classes.•Our outperforming results in terms of accuracy and recovering pathway.
论文关键词：Neural networks,Classification,Feature,Imbalanced data,Metabolite,Pathway
论文评审过程：Received 5 February 2017, Revised 14 June 2017, Accepted 15 June 2017, Available online 17 June 2017, Version of Record 3 July 2017.
论文官网地址：https://doi.org/10.1016/j.eswa.2017.06.026
