Detecting the Significant Flux Backbone of Escherichia coli metabolism

Abstract

The heterogeneity of computationally predicted reaction fluxes in metabolic networks within a single flux state can be exploited to detect their significant flux backbone. Here, we disclose the backbone of Escherichia coli, and compare it with the backbones of other bacteria. We find that, in general, the core of the backbones is mainly composed of reactions in energy metabolism corresponding to ancient pathways. In E. coli, the synthesis of nucleotides and the metabolism of lipids form smaller cores which rely critically on energy metabolism. Moreover, the consideration of different media leads to the identification of pathways sensitive to environmental changes. The metabolic backbone of an organism is thus useful to trace simultaneously both its evolution and adaptation fingerprints.Keywords: disparity filter; flux balance analysis; metabolic backbones; metabolic networks High-quality genome-scale metabolic reconstructions are composed of thousands of reactions and metabolites [1][2][3][4]. Due to their complexity, the analysis of these metabolic reconstructions requires computational approaches, like constraint-based optimization techniques [5,6], and methodological frameworks, like complex network science [7,8] Backbones maintain significant information while displaying a substantially decreased number of interconnections and, hence, can provide accurate but reduced versions of the whole system. In this direction, the work by Almaas et al.[13] introduced a filtering technique that selects the reactions dominating the production and consumption of each metabolite and connects two metabolites if the reaction producing one of them with the highest flux happens to be the reaction consuming the other with the highest flux, which defines a high-flux backbone. This method is able to segregate classical pathways, but the selected high-flux subgraphs present a

An evaluation tool for backbone extraction techniques in weighted complex networks

“…Backbone extraction has a wide range of applications. One uses it for various types of networks, such as social 68 – 76 , biological 77 , brain 78 – 81 , gene 82 – 86 , metabolic 87 – 91 , food web 92 , 93 , environmental 94 , 95 , finance 96 – 99 , trade 100 – 104 , information 105 – 108 , political 109 , 110 , transportation 111 – 115 , and others 116 – 119 . These applications have a broad range of uses, including clustering, classification, community detection, outlier detection, pattern set mining, identification of sources of infection in large graphs, and visualization, among others.…”

An Evaluation Tool for Backbone Extraction Techniques in Weighted Complex Networks

“…Backbone extraction has a wide range of applications. One uses it for various types of networks, such as social [61][62][63][64][65][66][67][68][69] , biological 70 , brain [71][72][73][74] , gene [75][76][77][78][79] , metabolic [80][81][82][83][84] , food web 85,86 , environmental 87,88 , finance [89][90][91][92] , trade [93][94][95][96][97] , information [98][99][100][101] , political 102,103 , transportation [104][105][106][107][108] , and others [109][110]…”