Limits...
Correlation-Based Network Analysis of Metabolite and Enzyme Profiles Reveals a Role of Citrate Biosynthesis in Modulating N and C Metabolism in Zea mays.

Toubiana D, Xue W, Zhang N, Kremling K, Gur A, Pilosof S, Gibon Y, Stitt M, Buckler ES, Fernie AR, Fait A - Front Plant Sci (2016)

Bottom Line: The overall higher CV values for metabolites as compared to the tested enzymes are indicative for their greater phenotypic plasticity.H(2) tests revealed galactinol (1) and asparagine (0.91) as the highest scorers among metabolites and nitrate reductase (0.73), NAD-glutamate dehydrogenase (0.52), and phosphoglucomutase (0.51) among enzymes.The latter displayed the strongest node-betweenness value (185.25) of all nodes highlighting its fundamental structural role in the connectivity of the network by linking between different communities and to the also strongly connected enzyme aldolase.

View Article: PubMed Central - PubMed

Affiliation: Institute of Dryland Biotechnology and Agriculture, Jacob Blaustein Institutes for Desert Research, Ben-Gurion University of the Negev Midreshet Ben-Gurion, Israel.

ABSTRACT
To investigate the natural variability of leaf metabolism and enzymatic activity in a maize inbred population, statistical and network analyses were employed on metabolite and enzyme profiles. The test of coefficient of variation showed that sugars and amino acids displayed opposite trends in their variance within the population, consistently with their related enzymes. The overall higher CV values for metabolites as compared to the tested enzymes are indicative for their greater phenotypic plasticity. H(2) tests revealed galactinol (1) and asparagine (0.91) as the highest scorers among metabolites and nitrate reductase (0.73), NAD-glutamate dehydrogenase (0.52), and phosphoglucomutase (0.51) among enzymes. The overall low H(2) scores for metabolites and enzymes are suggestive for a great environmental impact or gene-environment interaction. Correlation-based network generation followed by community detection analysis, partitioned the network into three main communities and one dyad, (i) reflecting the different levels of phenotypic plasticity of the two molecular classes as observed for the CV values and (ii) highlighting the concerted changes between classes of chemically related metabolites. Community 1 is composed mainly of enzymes and specialized metabolites, community 2' is enriched in N-containing compounds and phosphorylated-intermediates. The third community contains mainly organic acids and sugars. Cross-community linkages are supported by aspartate, by the photorespiration amino acids glycine and serine, by the metabolically related GABA and putrescine, and by citrate. The latter displayed the strongest node-betweenness value (185.25) of all nodes highlighting its fundamental structural role in the connectivity of the network by linking between different communities and to the also strongly connected enzyme aldolase.

No MeSH data available.


Correlation-based metabolite network with combined communities. Network visualization of metabolites as analyzed on the IBM population. Metabolites are presented as nodes and their relations as links. Computations of the correlations were conducted under the R environment. Cytoscape was used to generate graphical output of network. The Spearman rank correlation was employed to compute all pairwise correlations between metabolites across the entire set of inbred lines. Solely significant correlations are depicted (q ≤ 0.05 an r-value of ≥ 0.3). The relative width of edges corresponds to the absolute value of the estimated correlation coefficient. Positive correlations are denoted as blue edges, negative correlations are denoted as red edges. To distinguish between metabolites and enzymes in the network, nodes are represented in different shapes (metabolites = elliptical, enzymes = rectangular). Furthermore, nodes representing metabolites are color-coded according to their compound classes. The size of nodes corresponds to their relative connectedness (node degree). Metabolites are color-coded and clustered according to the walktrap community algorithm. The statistical significance of the occurrence of a community with more than four nodes was tested by performing a Wilcoxon signed rank test. The test was performed by assessing the degree of node-connectivity of the isolated community as compared to the nodes of the community still embedded in the network of which all community specific edges have been subtracted. Non-significant communities were combined with significant communities.
© Copyright Policy
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC4940414&req=5

Figure 4: Correlation-based metabolite network with combined communities. Network visualization of metabolites as analyzed on the IBM population. Metabolites are presented as nodes and their relations as links. Computations of the correlations were conducted under the R environment. Cytoscape was used to generate graphical output of network. The Spearman rank correlation was employed to compute all pairwise correlations between metabolites across the entire set of inbred lines. Solely significant correlations are depicted (q ≤ 0.05 an r-value of ≥ 0.3). The relative width of edges corresponds to the absolute value of the estimated correlation coefficient. Positive correlations are denoted as blue edges, negative correlations are denoted as red edges. To distinguish between metabolites and enzymes in the network, nodes are represented in different shapes (metabolites = elliptical, enzymes = rectangular). Furthermore, nodes representing metabolites are color-coded according to their compound classes. The size of nodes corresponds to their relative connectedness (node degree). Metabolites are color-coded and clustered according to the walktrap community algorithm. The statistical significance of the occurrence of a community with more than four nodes was tested by performing a Wilcoxon signed rank test. The test was performed by assessing the degree of node-connectivity of the isolated community as compared to the nodes of the community still embedded in the network of which all community specific edges have been subtracted. Non-significant communities were combined with significant communities.

Mentions: Here, we explore the relational ties between enzymes and metabolites implementing CNA to construct a unipartite network (Figure 4). Although the detected absolute correlation coefficients of the relationships between metabolites and enzymes ranged between values of 0.3 to 0.4—indicating rather moderate correlations—further investigations allowed us to suggest structures resembling known metabolic pathways.


Correlation-Based Network Analysis of Metabolite and Enzyme Profiles Reveals a Role of Citrate Biosynthesis in Modulating N and C Metabolism in Zea mays.

Toubiana D, Xue W, Zhang N, Kremling K, Gur A, Pilosof S, Gibon Y, Stitt M, Buckler ES, Fernie AR, Fait A - Front Plant Sci (2016)

Correlation-based metabolite network with combined communities. Network visualization of metabolites as analyzed on the IBM population. Metabolites are presented as nodes and their relations as links. Computations of the correlations were conducted under the R environment. Cytoscape was used to generate graphical output of network. The Spearman rank correlation was employed to compute all pairwise correlations between metabolites across the entire set of inbred lines. Solely significant correlations are depicted (q ≤ 0.05 an r-value of ≥ 0.3). The relative width of edges corresponds to the absolute value of the estimated correlation coefficient. Positive correlations are denoted as blue edges, negative correlations are denoted as red edges. To distinguish between metabolites and enzymes in the network, nodes are represented in different shapes (metabolites = elliptical, enzymes = rectangular). Furthermore, nodes representing metabolites are color-coded according to their compound classes. The size of nodes corresponds to their relative connectedness (node degree). Metabolites are color-coded and clustered according to the walktrap community algorithm. The statistical significance of the occurrence of a community with more than four nodes was tested by performing a Wilcoxon signed rank test. The test was performed by assessing the degree of node-connectivity of the isolated community as compared to the nodes of the community still embedded in the network of which all community specific edges have been subtracted. Non-significant communities were combined with significant communities.
© Copyright Policy
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC4940414&req=5

Figure 4: Correlation-based metabolite network with combined communities. Network visualization of metabolites as analyzed on the IBM population. Metabolites are presented as nodes and their relations as links. Computations of the correlations were conducted under the R environment. Cytoscape was used to generate graphical output of network. The Spearman rank correlation was employed to compute all pairwise correlations between metabolites across the entire set of inbred lines. Solely significant correlations are depicted (q ≤ 0.05 an r-value of ≥ 0.3). The relative width of edges corresponds to the absolute value of the estimated correlation coefficient. Positive correlations are denoted as blue edges, negative correlations are denoted as red edges. To distinguish between metabolites and enzymes in the network, nodes are represented in different shapes (metabolites = elliptical, enzymes = rectangular). Furthermore, nodes representing metabolites are color-coded according to their compound classes. The size of nodes corresponds to their relative connectedness (node degree). Metabolites are color-coded and clustered according to the walktrap community algorithm. The statistical significance of the occurrence of a community with more than four nodes was tested by performing a Wilcoxon signed rank test. The test was performed by assessing the degree of node-connectivity of the isolated community as compared to the nodes of the community still embedded in the network of which all community specific edges have been subtracted. Non-significant communities were combined with significant communities.
Mentions: Here, we explore the relational ties between enzymes and metabolites implementing CNA to construct a unipartite network (Figure 4). Although the detected absolute correlation coefficients of the relationships between metabolites and enzymes ranged between values of 0.3 to 0.4—indicating rather moderate correlations—further investigations allowed us to suggest structures resembling known metabolic pathways.

Bottom Line: The overall higher CV values for metabolites as compared to the tested enzymes are indicative for their greater phenotypic plasticity.H(2) tests revealed galactinol (1) and asparagine (0.91) as the highest scorers among metabolites and nitrate reductase (0.73), NAD-glutamate dehydrogenase (0.52), and phosphoglucomutase (0.51) among enzymes.The latter displayed the strongest node-betweenness value (185.25) of all nodes highlighting its fundamental structural role in the connectivity of the network by linking between different communities and to the also strongly connected enzyme aldolase.

View Article: PubMed Central - PubMed

Affiliation: Institute of Dryland Biotechnology and Agriculture, Jacob Blaustein Institutes for Desert Research, Ben-Gurion University of the Negev Midreshet Ben-Gurion, Israel.

ABSTRACT
To investigate the natural variability of leaf metabolism and enzymatic activity in a maize inbred population, statistical and network analyses were employed on metabolite and enzyme profiles. The test of coefficient of variation showed that sugars and amino acids displayed opposite trends in their variance within the population, consistently with their related enzymes. The overall higher CV values for metabolites as compared to the tested enzymes are indicative for their greater phenotypic plasticity. H(2) tests revealed galactinol (1) and asparagine (0.91) as the highest scorers among metabolites and nitrate reductase (0.73), NAD-glutamate dehydrogenase (0.52), and phosphoglucomutase (0.51) among enzymes. The overall low H(2) scores for metabolites and enzymes are suggestive for a great environmental impact or gene-environment interaction. Correlation-based network generation followed by community detection analysis, partitioned the network into three main communities and one dyad, (i) reflecting the different levels of phenotypic plasticity of the two molecular classes as observed for the CV values and (ii) highlighting the concerted changes between classes of chemically related metabolites. Community 1 is composed mainly of enzymes and specialized metabolites, community 2' is enriched in N-containing compounds and phosphorylated-intermediates. The third community contains mainly organic acids and sugars. Cross-community linkages are supported by aspartate, by the photorespiration amino acids glycine and serine, by the metabolically related GABA and putrescine, and by citrate. The latter displayed the strongest node-betweenness value (185.25) of all nodes highlighting its fundamental structural role in the connectivity of the network by linking between different communities and to the also strongly connected enzyme aldolase.

No MeSH data available.