What Is The Metabolic Hierarchy Animal Science

• Journal List
• PLoS One
• PMC5937743

PLoS 1. 2018; 13(v): e0195843.

A hierarchical model of metabolic machinery based on the kcore decomposition of constitute metabolic networks

Humberto A. Filho, Conceptualization, Data curation, Investigation, Methodology, Writing – original draft, Writing – review & editing, Jeaneth Machicao, Conceptualization, Investigation, Methodology, Visualization, Writing – original draft, Writing – review & editing, and Odemir Yard. Bruno, Supervision, Writing – review & editing ^*

Humberto A. Filho

São Carlos Found of Physics, University of São Paulo, São Carlos - SP, PO Box 369, 13560-970, Brazil

Jeaneth Machicao

São Carlos Plant of Physics, University of São Paulo, São Carlos - SP, PO Box 369, 13560-970, Brazil

Odemir M. Bruno

São Carlos Institute of Physics, University of São Paulo, São Carlos - SP, PO Box 369, 13560-970, Brazil

Diego F. Gomez-Casati, Editor

Received 2017 Sep fifteen; Accustomed 2018 Apr 1.

Supplementary Materials

S1 File: Properties from the 17 constitute metabolic networks. (PDF)

GUID: 945F0333-3CB9-4ADD-91D8-16303AC8E44B

S2 File: List of mutual metabolites. (XLSX)

GUID: 48AAB709-62FF-49B6-A50C-EDD9D3294F4D

S3 File: List secondary metabolites. (XLSX)

GUID: 2FD2EA7F-1781-42EA-B7F5-1A6A5FC1ADFA

S4 File: Listing of metabolites separated past classes or modules used for the cantankerous connectedness analysis. (XLSX)

GUID: 8EF9C5D8-E017-4D75-9DD1-21286B7B4C97

S5 File: List of metabolites per each layer. (XLSX)

GUID: 70B97175-D3EF-49DB-A0DC-3356600BB119

S6 File: Plant metabolic networks of the 17 plants species in GML format. (Zip)

GUID: FB7EBA22-3AF3-4CA2-9E73-AE3980CC1649

S7 File: Supplementary video. A video corresponding to the kcore decomposition for Arabidopsis thaliana metabolic network (Fig 1d). The video shows the decomposition process of the original network, layer past layer, until the maximum cadre is achieved. Each circle represents a metabolic node. The colors represent the layer for which the nodes vest to.

(MOV)

GUID: C9F334DD-F0A6-48B8-8A53-6BF5327FA7DF

Data Availability Argument

All relevant data are within the paper and its Supporting Data files.

Abstract

Modeling the basic structure of metabolic machinery is a challenge for modern biology. Some models based on complex networks accept provided important information regarding this mechanism. In this newspaper, we synthetic metabolic networks of 17 plants covering unicellular organisms to more circuitous dicotyledonous plants. The metabolic networks were built based on the substrate-production model and a topological percolation was performed using the mcadre decomposition. The distribution of metabolites across the percolation layers showed correlations between the metabolic integration hierarchy and the network topology. We testify that metabolites concentrated in the internal network (maximum thoucadre) only comprise molecules of the principal basal metabolism. Moreover, we institute a high proportion of a set of mutual metabolites, among the 17 plants, centered at the inner kcore layers. Meanwhile, the metabolites recognized every bit participants in the secondary metabolism of plants are full-bodied in the outermost layers of the network. This data suggests that the metabolites in the central layer form a bones molecular module in which the whole plant metabolism is anchored. The elements from this fundamental core participate in nigh all institute metabolic reactions, which suggests that plant metabolic networks follows a centralized topology.

Introduction

It is believed that the general architecture of living systems is based on the structure and function of modules [1]. However, information technology is still a claiming for biology to empathise how this modular organization is structured from the molecular to the cellular level. A model that relates how biological modules collaborate would be very useful. The metabolism of living beings has been understood as a network of interactions that has a modular organization [two]. At the aforementioned fourth dimension the understanding of this metabolism as a complex network has brought new visions virtually the organization of the metabolic components and the network topology of possible interactions between the metabolites [three].

The metabolism is perchance the all-time network of interactions e'er characterized in biology because a large number of studies has defined metabolic pathways [four, 5]. For subsequent decades of enzymology applications, the catalytic and regulatory properties of enzymes take been characterized. More recently genetic studies and molecular biology have opened upwards avenues in the knowledge of enzyme genes that catalyze reactions of transformations of affair in living beings. Thus, there is an unprecedented amount of descriptive and mechanistic information on the behavior of metabolic phenomena that tin exist analyzed as complex networks of which relevant information about living systems tin be obtained.

Data extracted from constitute genomes make information technology possible to investigate correlations of the molecular courage of living beings that are related to their adaptations to the environment [half-dozen–8]. Complete information from the genome of diverse plants take generated a set of metabolic reactions based on the genes of enzymes that can resume the metabolism of these organisms. The metabolism is divided into detached pathways, however it operates equally a highly integrated network. Therefore, metabolic networks can be built and the metabolism of various living beings is modeled and quantified based on network parameters [3].

However, it is non known exactly how a general design on the distribution of matter in metabolism can be developed based on the observation of the topology from metabolic networks. Consequently, in social club to reveal functional, hierarchical and fifty-fifty phylogenetic information hidden in the network structure, new models demand to be investigated. The content of found metabolic networks reflects the species phylogeny, thus groups of plants that are evolutionary closer share a larger proportion of mutual metabolites. In this instance, the specialized metabolism is responsible for the greatest differences; gene coding for specialized metabolic functions have proliferated to a much greater caste and by different mechanisms and display lineage-specific patterns of physical clustering inside the genome [9]. Thus, understanding the connectivity blueprint of secondary metabolites in the metabolic networks of plants should be central to the overall understanding of plant metabolism.

Although some studies have shown that the hierarchical arrangement and modularity within the metabolic network is related to the chemic classes of metabolites [2], these studies lack a holistic model concerning how the arrangement of modules is centralized. This comprehensive model would find general principles that govern the structure and function of modules.

For case, some attempts accept been made to find fully-connected parts of the metabolic networks [10]. Completely continued cores of networks, in general, have been obtained past the kcore decomposition [11, 12]. It is a well-known algorithm that has been used in many contexts, mostly in biological networks, such as poly peptide-protein interaction [13]. Through kcore percolation, a network is pealed every bit a set of successively mcores (layers (chiliad)) until a maximum cadre is achieved. Thus, modest values of yard define the periphery of the network and the innermost network core corresponds to a big thousand. This algorithm has been proposed to understand the network hierarchical arrangement past extracting highly interconnected parts and thus providing ways to observe relationships within these substructures [14]. Therefore, the kcores can provide useful insights into the global network topology and too establish a hierarchy of connections from each network node [15].

In this report, we present a percolation analysis based on the kcore decomposition of plant metabolic networks modeled according to the institute metabolic reaction dataset from 17 plants selected from the PlantCyc database [16]. The metabolic networks were built by using a well-known model "substrate-product network" [17], where metabolites (nodes) from substrates are linked to metabolites from products of the reactions. Thus, the metabolic network of 17 plants, namely: Brachypodium distachyon (BD), Hordeum vulgare (HV), Oryza sativa japonica (OSJ), Panicum virgatum (PV), Setaria italica (SI), Sorghum bicolor (SB), Zea mays (ZM); Arabidopsis thaliana (AT), Brassica rapa pekinensis (BRP), Carica papaya (CP), Glycine max (GM), Manihot esculenta (ME), Populus trichocarpa (PT), Vitis vinifera (VV); Selaginella moellendorffii (SM); Physcomitrella patens (PP) and Chlamydomonas reinhardtii (CR) were considered.

Our primary objective is to demonstrate that a percolation through the grandcore decomposition of plant metabolic networks reveals data about the metabolic centralization and functional hierarchy of certain metabolite groups distributed across its layers.

This functional analysis shows that a high proportion of the metabolites from cardinal layers of the networks contain the basic components from basal metabolism. In addition, we bear witness that the metabolites found in the more centralized level of the networks integrate all plant metabolisms considering they participate in the absolute bulk of their metabolic reactions. Every bit confirmed by previous studies that show that metabolic networks are very modularized [two], we besides find that institute metabolic networks form modules. The metabolites contained in these modules connect to the metabolites found in the fundamental core of the networks in a much larger proportion than the modules themselves, which suggests that the primal core can course a kind of metabolic module of the metabolic networks and that there is sub-modularization in the networks as has already been suggested in previous studies [2]. In addition, we establish that basal metabolism is distributed along the gcore layers of the networks in an reverse mode relative to the secondary and non-common metabolism between plants. The common metabolism between plants is completely centralized in the layers with greater connectivity and the metabolism secondary and uncommon between the plants is peripheral forth the layers. Together these data prove that the decomposition by a layering tool is useful for building hierarchical models of the metabolism arrangement and brings information that can be used in studies about the development of the metabolism.

Results

kcore percolation of plant metabolic networks

We practical the grandcore percolation algorithm to 17 plant metabolic networks whose gear up of metabolic reactions were obtained from the PlantCyc database [sixteen]. The fix of metabolic reactions from each plant was used to congenital metabolic networks following the substrate-product model [17], which is a model that establishes the correlation of distribution of matter in metabolisms through a network of connections related to the formation and decomposition of components [xviii].

The Fig 1a shows a given gear up of reactions from which a network is obtained. The kcore decomposition is applied successively, as shown in Fig 1b, until a maximum core is obtained, whose corresponding adjacency matrix is shown in Fig 1c. The overall sketching of the distribution of layers in a generic metabolic network is shown in Fig 1d.

(a) Example of a prepare of simplified metabolic reactions that assemble a metabolic network, which is built because each molecule of the reaction according to the substrate-product network model [17]. Thus, each substrate metabolite is individually linked to each one of the product metabolites and vice versa. (b) The 1000core decomposition of a given network. First iteration: the peripheral nodes with degree k _i = one (dark gray) are removed. Second iteration: The remaining network of the start iteration, which presents 9 nodes of which 2 of them contain only one border (low-cal gray), is pruned once again. Finally, the resulting 2-cadre corresponds to a fully connected cadre composed of 7 nodes (white). (c) Resulting adjacency matrix corresponding to the ii-core crush where an approachable link from the nodes in the outset row of the matrix has an incoming link at each metabolite from the get-go column, and vice versa. (d) kcore decomposition from Arabidopsis thaliana metabolic network. From top to downwardly, the original network at 1000 _one is pealed iteratively until thou _max where it reaches k = 18 layers (see S7 File). Different colors were used to represent each layer. Edges are not shown.

The one thousandcore percolation from the metabolic network of Arabidopsis thaliana shows that the concluding full continued level (maximum thousandcore) contains 21 metabolites, while the outermost layer contains 3,546 metabolites, from which 149 belong exclusively to k ₁. As expected, at that place is a subtract in the number of metabolites in each layer until the cardinal core. It tin be observed that the maximum number of layers k _max for all the studied plants varies between 16 and 18. Thus, these plants were grouped into three sets: BD, HV, OSJ, SI and SB with sixteen layers; PV, ZM, BRP, CP, GM, ME, VV, SM and CR with 17 layers; AT and PP with 18 layers.

We institute that the number of layers thousand _max is not related to the number of metabolites (N). For case, networks with relatively small sizes such equally moss Physicomitrela patens (North = 2651) reached the highest number of cores (chiliad = 18), while networks with larger sizes such as the dicotyledonous Manihot esculenta (N = 2991) reached fewer cores (k = 16). The detailed network backdrop, such as the number of layers and network size for the 17 constitute metabolic networks studied here, can exist establish in S1 File.

We as well analyzed the chemic composition of the fundamental core in order to obtain the contents of the innermost network for the 17 plants, which is shown in Fig 2. Information technology was observed that the central cadre of the constitute networks only comprises molecules from the basal metabolism of plants. In this regard, 14 metabolites were found in common at the maximum kcore from all plants studied hither namely: ADP, ammonia, AMP, ATP, CO₂, coenzyme A, diphosphate, H+, H₂O, L-glutamate, NAD+, NADH, NADPH, and phosphate. These 14 metabolites extracted from the primal core, institute in the 17 studied plants, represent a very connected dataset.

Metabolites institute at the maximum cadrechiliad _max.

The columns indicate the plant acronyms, while the metabolites are shown as rows. The white cells point the presence of a metabolite in a particular plant chiliadcadre, while black cells indicate its absenteeism.

Integration of metabolism within the key yardcore

We observed that the metabolites extracted from the key cadre of the metabolic networks participate in most all the sets of the metabolic reactions from plants studied hither. This is shown in Fig 3, where it tin be observed that the proportion of metabolic reactions containing at least 1 metabolite belonging to the central core of each institute is much higher than the number of reactions that do non present it. Therefore, the results testify that all the metabolic pathways are integrated with the most basal metabolism.

Found reactions that contain at to the lowest degree one metabolite from the maximum kcore shown by a bluish bar, otherwise by a grayness bar.

In earlier studies, it was revealed that metabolic networks of living beings accept a modular and hierarchical structure [2]. In lodge to assess the levels of modularization of plant metabolic networks, we evaluated the connectivities of metabolic classes (modules) within the maximum kcore to the metabolic network from Arabidopsis thaliana (see Fig four). Nodes contained in the central layer that belong to any one of the addressed classes were removed from their respective classes and grouped in the maximum kcore class in order to avert ambiguity between the classes (encounter Material and methods section).

(a)-(e) Cross connection of metabolites from Arabidopsis thaliana clustered past metabolic modules: Lipids, Carbohydrates, kcore fundamental (cadre), Nucleotides and Amino acids. Each plot shows the number of links from metabolites, belonging to a respective module that are continued to metabolites from other modules. (f) Proportion of metabolites per modules relative to the full content of metabolites analyzed.

The graphs in Fig four demonstrate the high level of modularization of the metabolic network of Arabidopsis thaliana and also the high connectivity of the modules (metabolic classes) with the maximum grandcadre of the network. It tin exist seen for example in Fig 4a that the lipids connect to other lipids and too to the primal core in a much higher proportion than the other metabolic classes. It tin also be observed in Fig 4c that the metabolites contained in the most central layer of the network are connected to the other metabolic classes at a higher proportion in all classes, which shows the enormous connectivity with the central layer. Therefore, it is clear that the network is modularized and centralizes the connections of the modules with the almost central layer (maximum thoucore).

There are differences in the number of the metabolites in each form shown in Fig 4a to 4e. The proportion of metabolites from each class relative to the full content of metabolites in all classes is demonstrated in Fig 4f. The number of connections from each class is absolute, therefore the level of centralization in the key core is very high. For example the maximum grandcore form contains merely 21 metabolites while the lipids incorporate effectually 500. This is about 20 times higher than the content of the metabolites from the maximum yardcore, while the number of the links at the maximum yardcore is almost the same every bit the lipids.

Differential distribution of the metabolism over the thousandcore percolation layers

We analyzed the proportion of common metabolites against the full content of metabolites across each chiliadlayer. These metabolites were extracted from the intersection dataset betwixt the 17 plants studied in this piece of work. The Fig 5a, shows the distribution of common metabolites across the klayers for plants belonging to five major clades, namely: the algae Chlamydomonas reinhardtii, the bryophyta moss Physicomitrela patens, the lycophyte Selaginella moelendorffi and two angiosperms, the monocotyledon Zea mays and the dicotyledon Arabidopsis thaliana. The analysis of the proportion from common metabolites in the kcore layers reveals that its content is centered in the innermost one thousandcore, where these metabolites reach to nigh the total content of the layers. The distribution of the metabolites per each kcore layer for plants that were not shown in Fig v can be found in S1 File. On the other manus, the distribution of the complementary set to the common metabolism, the not-mutual metabolites betwixt 1 plant and the other sixteen plants studied, has an contrary decaying blueprint along the kcore layers, every bit it is a complementary set of the common metabolites, as shown in Fig 5b. Moreover, the previous analysis was repeated also to clarify the content of the secondary metabolites (run across Fig 5c). Interestingly this distribution presented the same decaying pattern over the klayers equally that shown to not-common metabolites between the plants (see Fig 5b).

(a) Pct of common metabolites shared with full content from each kcore distributed along the percolation layers of the plant metabolic networks from Chlamydomonas reinhardtii, (CR) Physicomitrela patens, (PP) Selaginella moelendorffi, (SM) Zea mays(ZM) and Arabidopsis thaliana(AT). (b) Percent of non-common metabolites distribution along the thoulayers. (c) Proportion of secondary metabolites shared with full content of the metabolites from each kcore percolation level from the same 5 plants shown in (a).

This result reveals that the secondary metabolites are concentrated in the periphery of the networks and are absent in the inner layers. The distribution of the content of secondary metabolites along the mcore layers is completely related to the not-common metabolic content betwixt the plants. It is, therefore, possible that the distribution of the non-mutual metabolism throughout the thousandlayers can exist estimated equally a reflection of the distribution of secondary metabolism.

At the same time, this analysis shows that the database of secondary metabolites, specifically cured for this work, tin can be estimated as a significant sample of the secondary metabolite content independent in PlantCyc. Its distribution over the grandlayers shows that the topological analysis of the network, by means of the thoucore algorithm, reveals data that correlates the topological construction of the metabolic network with the metabolic functionality, therefore discerning the basal and secondary metabolism co-ordinate to their different topologies. Although algae, bryophytes and lycophytes bear witness a much smaller amount of secondary metabolites in relation to the angiosperms, we observed that in the species of these clades, the secondary metabolites also take a peripheral location across the kcores.

It is important to emphasize that there are some secondary metabolites contained inside the dataset of mutual metabolites establish in the 17 plants. However, they are the minority and the assay of the distribution of the non-common metabolites excludes the possibility of bias in our analysis of the distribution of the differential disuse of the metabolism over the glayer. At the aforementioned fourth dimension, it is interesting to notice that the simpler organism studied hither is a unicellular algae Chlamydomonas reinhardtii, which already provides specialization of its metabolism and some secondary metabolites. Therefore, the intersection between the secondary and common metabolites amidst the 17 plants studied here probably derives from the fact that unicellular algae already present metabolic specialization. Additionally, information technology is likewise interesting to detect that the amount of secondary metabolites establish at the dicotyledon Arabidopsis thaliana is greater than the ones found in the other plants (run into Material and methods department and S1 File). Of course, as it is considered a model constitute for the study of angiosperms, its molecular biological science and biochemistry have been exhaustively studied, which logically encompasses its metabolism. This fact is clearly evidenced by the number of reactions available at the PlantCyc database. The results show that the bones general limerick of the metabolism of whatever institute is centered in the innermost kcore percolation layers and that metabolites belonging to these inner layers, with minor variations amid the plants, are completely contained in the set of mutual metabolites of all plants. This also shows that the secondary metabolism is full-bodied in the outer layers.

From Fig 5a and 5c, we can clearly observe an contrary relationship between secondary and mutual metabolites regarding the distribution of metabolites across the layers of the networks. This suggests that there may be a correlation between the metabolism specialization and the hierarchy established artificially past percolation in the kcadre layers.

This issue supports the hypothesis that percolation in kcores provides information about the metabolism specialization and distribution of metabolites in circuitous networks associated with the topology. It besides suggests the thought that the molecular evolution of metabolism can also exist modeled by percolation in kcores.

It is known that the metabolism specialization forth the institute evolution follows a design of descent with modification, with closely related species sharing more like sets of metabolic reactions [nine]. Therefore, it is also possible that the 1000cadre percolation reveals evolutionary information between institute species.

Discussion

In this newspaper, we presented a model of the hierarchical construction of plant metabolic networks by means of the thoucore percolation layers, from which we observed the cardinal cadre of 17 establish metabolic networks, reached past percolation. The metabolites found in this core were mainly the about bones components of the constitute basal metabolism such equally some ions used as electron transport (H ⁺, P O four 3 - , H C O 3 - ), metabolic currency (ATP, ADP, AMP), electron carriers (NADP+, NADPH, NAD, NADH), H _two O, metabolites of the most basic energetic metabolism (pyruvate, oxaloacetate, Acetyl-Coa), amino acids (L-glutamate) and the most primary photosynthesis production, the oxygen (O ₂) (see Fig ii). This fact suggests that percolation generates some hierarchy of metabolic complexity along the germination of layers in the network.

The metabolite distribution across the establish'southward metabolic reactions showed that at least one chemical element of the set up of metabolites belonging to the most central layer of the networks is present in more than 90% of the metabolic reactions from all plants studied here (encounter Fig three). It is suggested that the key cadre elements tin grade a module that integrate all plant metabolic networks.

The analysis of the cross connexion between metabolic classes showed that the metabolic classes in the Arabidopsis thaliana are highly continued within themselves, as the number of connections inside each metabolic class (Lipids, Carbohydrates, Nucleotides, and Amino acids) is higher than with other classes. Interestingly all classes are very connected with metabolites nowadays at the maximum kcore layer. This shows that in addition to beingness very modularized, the network is likewise very centralized in the metabolites found in the last level of the network percolation. Earlier studies have demonstrated that the metabolic networks accept a modular and hierarchical structure [two].

It is also interesting to note that the common metabolism, represented by mutual metabolites betwixt all plants studied here, is centralized in the mcores. Moreover, it is known that the basal metabolism arose early in the development of species, therefore it is not unreasonable to assume that the percolation by kcores provides a tool for studying molecular evolution metabolism, which is supposed to occur with the direction of the center to the periphery of the network percolation layers.

The peripheral location of the secondary and not-mutual metabolites between plants forth the kcadre layers also corroborates the hypothesis that the kcore percolation algorithm may reveal evolutionary hierarchies in the metabolism and information technology is appropriate to demonstrate connectivity hierarchies in plant metabolic networks and their correlation with the metabolic functionality.

It should be noted that the secondary metabolism provides alternative ways of molecular interactions between the plant and its environment and for this reason it appeared subsequently in the evolution of species. This metabolism has a differential evolution in relation to the basal metabolism and genes related to information technology proliferate and diverge more rapidly throughout the evolution [9].

Regarding the curated dataset of secondary metabolites detailed in Material and Methods department, we must clarify that in that location is not an absolute distinction betwixt basal and secondary metabolites, however, in general, information technology is meant by basal metabolism the prepare of metabolic reactions and metabolites that class the basic framework of metabolic pathways in mutual among all plants, from algae to dicotyledons. As well the list of curated secondary metabolites, nosotros have chosen to bring an additional data which refers to the distribution of non-common metabolites betwixt plants inside the framework of the kcadre percolation. Nosotros observed that the distribution of these metabolites is very like to that of the secondary metabolites (see Fig 5). This fact was predictable and shows that the not-common metabolism between plants reflects the secondary metabolism. At the same time, this effigy shows that, in fact, the secondary metabolism has a peripheral distribution in the layers of the network. In addition, it shows that the information generated in the database is pertinent and reveals topological information about plant metabolic networks.

The topological structure of the metabolic pathways forth the kcore layers tin be assembled as a huge puzzle. Therefore, the logic of the formation of pathways may be elucidated through tracking pathway reactions forth the kcore percolation layers in different constitute networks. Based on this, a new question can be put forward: How are the parts of the paths grouped to grade a complete path along the gcores? The reply to this puzzle may simultaneously be how the molecular evolution of these routes is built in the various varieties of plants.

The kcore percolation can also be useful to explicate the operation of the routes in terms of the centralization of distribution of cloth (metabolites) forth the topology of the networks, every bit the vast majority of the reactions comprise at least ane element of the cardinal layer and therefore, announced to be embedded in the central module. This suggests that ship of matter at the metabolic pathways does non have an occasional structure and information technology can be centered in terms of routes. Moreover, it tin be assumed that send is performed by a chemical network, which although is not spatially defined, has logical connections that distribute material components (molecules). Therefore, theoretical correlations about the genesis and dynamics of pathways tin be drawn from the knowledge of key metabolic nodes, which could be shown by metabolomic and transcriptomic data, for instance. Therefore the grandcadre analysis can reveal the hierarchy and emergence of these nodes in various topological states of the metabolic network.

All elements of the key core are considered hubs in plant metabolic networks. These metabolites are the bones bricks of which all metabolic pathways are built. Meanwhile, nosotros observed that the metabolites at the maximum kcore also participate in almost all chemical reactions of the secondary metabolism. This observation leads to proposing a model in which the one thousandcore central layer is a kind of network processing unit of measurement, which is responsible for anchoring the constitute metabolic reactions.

Hierarchical network models evidence that their assembly hierarchy is perhaps associated with the specialization of metabolism [ii]. At the same time, our results reveal that the basal metabolism has strongly connected network nodes (metabolites) and the secondary metabolism has a more than peripheral location beyond the yardcore percolation layers. It suggests that studies on the modularization of metabolism could be performed with mcadre percolation to reveals hierarchies of modularity in metabolism.

Finally, some questions remain: Can we assume that the molecular evolution of metabolism, i.e. adding and integrating new components tin be simulated by the percolation proposed in this study? Is this addition anchored to the metabolites of the central layers?

Material and methods

Metabolic reaction database

The constitute metabolic networks studied here were constructed from the metabolic reactions of 17 plants available at PlantCyc database (Version 9.v) [16], further details on how this database was constructed tin be obtained in Ref. [xix]. This database contains the reconstruction of the found biochemical pathways based on an enzyme sequence from annotated metabolic functions of protein sequences [19–24]. This cognition has allowed researchers to found databases that are a skilful approximation of the content of plant metabolism and metabolic pathways. Although information technology tin can not be stated that the content of the reactions available in PlantCyc database is complete, it certainly reflects the annotation of much of the universe of known metabolic reactions and of their conference according to the annotated reactions based on gene sequences from enzymes. Certainly, until now, it is non known how many metabolic reactions can still be plant at whatsoever plant through metabolomic science [25]. Withal metabolic reactions from plants with genome sequenced have been annotated based on DNA sequence information from known enzymes assigned according to reactions discovered. More than details virtually how shut are the information from metabolic reaction datasets relative to the complete annotation of metabolomes can be found in Ref. [25].

Table 1 shows the corresponding web references of the reaction datasets used in this piece of work. This dataset provides information for the development of new hypotheses about the evolution and specialization of the constitute metabolism [9]. Some hypotheses created from the bachelor information were also used to develop our model.

Table 1

List of websites for metabolic reaction sets of the 17 plants studied in this work.

These lists tin be plant in the Plantcyc database (version 9.5) [16].

Datasets

• Mutual metabolites This dataset is composed by the intersection among all metabolites of plants analyzed here. The dataset is bachelor in S2 File.

• Non-mutual metabolites This dataset is complementary to the one described above, i.eastward. it is equanimous past those elements that exercise not belong to the intersection of metabolites among all the plants studied.

• Secondary metabolites This dataset was obtained by discriminating the complete dataset of metabolites of each establish. The secondary metabolites were identified using well-established literature regarding secondary metabolism [26]. The number of curated secondary metabolites per plant is shown in Tabular array two. The PlantCyc compound database [16] and the PubChem research tool [27] were used as a complementary resource to curate the listing of secondary metabolites by comparison each metabolite confronting these databases. Although these databases may exist partial in the classification between secondary and basal metabolism, their information is complementary (data available in S3 File).

  Table 2

  Number of secondary metabolites per plant.

  Plant	num. metabolites	Institute	#	Plant	#
  BD	633	ZM	668	PT	732
  HV	619	AT	824	VV	695
  OSJ	691	BRP	750	SM	476
  PV	690	CP	721	PP	373
  SI	648	GM	706	CR	163
  SB	645	ME	706

• Metabolic classes This data set was obtained from the attribution of metabolites belonging to the classes: amino acids, nucleotides, lipids, carbohydrates and those metabolites found in the maximum kcore of the metabolic network of the plant Arabidopsis thaliana (data available in S4 File). The metabolites from the maximum thousandcore module that eventually belongs to the other modules were withdrawn from this course and reclassified as the maximum 1000core metabolite to avoid ambiguity.

• Metabolites per kcore layer The metabolite content for each plant metabolic network, distributed across thoucore percolation layers, was obtained by listing the metabolites during each level of the network percolation with the kcore algorithm (information bachelor in S5 File).

Metabolic network construction

We considered metabolic networks following the substrate-product network model [17]. The gear up of metabolic reactions from each plant was reduced into a metabolic network represented by an undirected graph consisting of a set of North metabolites.

The establish metabolic networks were modeled using the fix of biochemical reactions from a specific plant. The set of metabolic networks used here are compiled and bachelor in S6 File. Fig 1a and 1b) shows an example of the metabolic network modeling. For the sake of analogy, the plot shows two metabolic reactions that correspond to the network modeling in Fig 1b. In order to build the metabolic network, each metabolite reaction was considered. Thus, each metabolite is individually linked to each other with respect to the reaction arrows. Therefore, repeated metabolites and/or edges are dismissed. The metabolic network is represented by the adjacency matrix A of Northward × N, where each element a _ij is related with each other by a link.

kcadre decomposition algorithm

The kcore decomposition is an algorithm that splits a network into hierarchically ordered sub-structures (run into S7 File). A kcore layer is the maximal subgraph obtained past recursively removing all nodes with a caste lower than k until all nodes in the remaining graph have a caste larger than or equal to one thousand [11, 12]. As this algorithm is an iterative procedure, it should non be confused with pruning nodes of a certain degree [15]. Fig 1b shows the topological percolation process in which the kcores are obtained from a given network. The node i in this network has a layer index k if it belongs to the thousandcore but not to the (k + ane)-core. A one thousandshell G _thousand is composed past all the nodes whose crush index is k [15]. The maximum value one thousand in a given 1000 _k is denoted by yard _max [15].

Cantankerous connexion of the metabolic modules

Metabolites belonging to the classes amino acids, nucleotides, lipids, carbohydrates and maximum chiliadcadre were discriminated by identifying metabolites from each corresponding class nowadays within the full content of metabolites from Arabidopsis thaliana.

The quantification of the cantankerous connection of metabolites was performed by counting the number of links betwixt metabolites belonging to a specific module with metabolites from other modules.

Supporting data

S1 File

Backdrop from the 17 plant metabolic networks.

(PDF)

S2 File

List of common metabolites.

(XLSX)

S3 File

List secondary metabolites.

(XLSX)

S4 File

List of metabolites separated by classes or modules used for the cross connection analysis.

(XLSX)

S5 File

List of metabolites per each layer.

(XLSX)

S6 File

Plant metabolic networks of the 17 plants species in GML format.

(ZIP)

S7 File

Supplementary video.

A video corresponding to the chiliadcadre decomposition for Arabidopsis thaliana metabolic network (Fig 1d). The video shows the decomposition process of the original network, layer by layer, until the maximum core is achieved. Each circle represents a metabolic node. The colors represent the layer for which the nodes belong to.

(MOV)

Acknowledgments

H. A. F. gratefully acknowledges the fiscal support of CNPq (National Council for Scientific and Technological Development, Brazil) (grant #153137/2013-four). J.K. is grateful for the support of CAPES (Coordination for the Improvement of College Education Personnel) and CNPq grant #405503/2017-two. O.M.B. gratefully acknowledges the financial support of CNPq (grants #307797/2014-seven, #405503/2017-ii and #153137/2013-4) and FAPESP (São Paulo Research Foundation) (grant #2014/08026-i). The funders had no office in report blueprint, information collection and analysis, determination to publish, or training of the manuscript.

Funding Statement

H. A. F. gratefully acknowledges the financial back up of CNPq (National Council for Scientific and Technological Development, Brazil) (grant #153137/2013-4). J.G. is grateful for the support of CAPES (Coordination for the Improvement of Higher Educational activity Personnel) and CNPq grant #405503/2017-2. O.M.B. gratefully acknowledges the fiscal support of CNPq (grants #307797/2014-seven, #405503/2017-2 and #153137/2013-iv) and FAPESP (São Paulo Research Foundation) (grant #2014/08026-ane). The funders had no role in report design, data drove and analysis, decision to publish, or preparation of the manuscript.

Data Availability

All relevant data are within the paper and its Supporting Data files.

References

1. Hartwell LH, Hopfield JJ, Leibler S, Murray AW. From molecular to modular jail cell biology. Nature. 1999;402:47–52. doi: 10.1038/35011540 [PubMed] [Google Scholar]

2. Ravasz Eastward, Somera AL, Mongru DA, Oltvai ZN, Barabási AL. Hierarchical organization of modularity in metabolic networks. Science. 2002;297:1551–1555. doi: 10.1126/science.1073374 [PubMed] [Google Scholar]

iii. Jeong H, Tombor B, Albert R, Oltvai ZN, Barabási AL. The big-scale system of metabolic networks. Nature. 2000;407(6804):651–654. doi: 10.1038/35036627 [PubMed] [Google Scholar]

four. Bassham JA, Calvin K. The path of carbon in photosynthesis In: Die CO2-Assimilation/The Absorption of Carbon Dioxide. Springer; 1960. p. 884–922. [Google Scholar]

five. Kornberg H. Krebs and his trinity of cycles. Nature Reviews Molecular Cell Biology. 2000;one(iii):225–228. doi: ten.1038/35043073 [PubMed] [Google Scholar]

vi. Derelle E, Ferraz C, Rombauts Southward, Rouze P, Worden AZ, Robbens South, et al. Genome analysis of the smallest free-living eukaryote Ostreococcus tauri unveils many unique features. Proceedings of the National Academy of Sciences. 2006;31:11647–11652. doi: 10.1073/pnas.0604795103 [PMC free article] [PubMed] [Google Scholar]

7. Merchant SS, Prochnik SE, Vallon O, Harris EH, Karpowicz SJ, Witman GB, et al. The Chlamydomonas genome reveals the evolution of key brute and constitute functions. Science. 2007;318:245–250. doi: ten.1126/science.1143609 [PMC gratis article] [PubMed] [Google Scholar]

8. Initiative TAG. Assay of the genome sequence of the flowering constitute Arabidopsis thaliana. Nature. 2000;408:796–815. doi: 10.1038/35048692 [PubMed] [Google Scholar]

nine. Chae L, Kim T, Nilo-Poyanco R, Rhee SY. Genomic signatures of specialized metabolism in plants. Science. 2014;344:510–513. doi: x.1126/science.1252076 [PubMed] [Google Scholar]

ten. Ma HW, Zeng AP. The connectivity construction, giant strong component and centrality of metabolic networks. Bioinformatics. 2003;19(xi):1423–1430. doi: 10.1093/bioinformatics/btg177 [PubMed] [Google Scholar]

11. Batagelj V, Zaversnik M. Generalized Cores. CoRR. 2002;cs.DS/0202039.

15. Alvarez-Hamelin JI, Dall'Asta 50, Barrat A, Vespignani A. Big scale networks fingerprinting and visualization using the kcadre decomposition In: Y Weiss BS, Platt J, editors. Advances in neural information processing systems. Cambridge (Massachusetts): MIT Press; 2006. p. 41–51. [Google Scholar]

17. Zhao J, Yu H, Luo J, Cao ZW, Li Y. Complex networks theory for analyzing metabolic networks. Chinese Science Bulletin. 2006;51(xiii):1529–1537. doi: x.1007/s11434-006-2015-2 [Google Scholar]

xviii. Holme P, Huss M. Substance graphs are optimal elementary-graph representations of metabolism. Chinese Science Message. 2010;55(27):3161–3168. doi: ten.1007/s11434-010-4086-three [Google Scholar]

nineteen. Zhang P, Dreher Thousand, Karthikeyan A, Chi A, Pujar A, Caspi R, et al. Cosmos of a genome-wide metabolic pathway database for Populus trichocarpa using a new approach for reconstruction and curation of metabolic pathways for plants. Plant Physiology. 2010;153(four):1479–1491. doi: x.1104/pp.110.157396 [PMC free article] [PubMed] [Google Scholar]

21. Rhee SY, Zhang P, Foerster H, Tissier C. AraCyc: overview of an Arabidopsis metabolism database and its applications for plant research. Plant Metabolomics. 2006; p. 141–154. doi: 10.1007/3-540-29782-0_11 [Google Scholar]

22. Dreher K. Putting The Plant Metabolic Network pathway databases to work: going offline to proceeds new capabilities. Constitute Metabolism: Methods and Protocols. 2014; p. 151–171. doi: ten.1007/978-1-62703-661-0_10 [PubMed] [Google Scholar]

23. Zhang P, Foerster H, Tissier CP, Mueller Fifty, Paley S, Karp PD, et al. MetaCyc and AraCyc. Metabolic pathway databases for constitute research. Plant physiology. 2005;138(ane):27–37. doi: x.1104/pp.105.060376 [PMC free article] [PubMed] [Google Scholar]

24. Caspi R, Foerster H, Fulcher CA, Kaipa P, Krummenacker M, Latendresse M, et al. The MetaCyc Database of metabolic pathways and enzymes and the BioCyc drove of Pathway/Genome Databases. Nucleic acids enquiry. 2007;36(suppl_1):D623–D631. doi: x.1093/nar/gkm900 [PMC gratis commodity] [PubMed] [Google Scholar]

26. Seigler DS. Plant Secondary Metabolism. Springer Scientific discipline & Business Media; 2012. [Google Scholar]

27. Kim Southward, Thiessen PA, Bolton EE, Chen J, Fu G, Gindulyte A, et al. PubChem substance and chemical compound databases. Nucleic acids inquiry. 2015;44(D1):D1202–D1213. doi: 10.1093/nar/gkv951 [PMC costless article] [PubMed] [Google Scholar]

---

Articles from PLoS ONE are provided here courtesy of Public Library of Scientific discipline

---

Source: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5937743/

Posted by: hornupout1979.blogspot.com

Share this post