Real Biological Networks Research Articles

Theory of interface: Category theory, directed networks and evolution of biological networks

Biological networks have two modes. The first mode is static: a network is a passage on which something flows. The second mode is dynamic: a network is a pattern constructed by gluing functions of entities constituting the network. In this paper, first we discuss that these two modes can be associated with the category theoretic duality (adjunction) and derive a natural network structure (a path notion) for each mode by appealing to the category theoretic universality. The path notion corresponding to the static mode is just the usual directed path. The path notion for the dynamic mode is called lateral path which is the alternating path considered on the set of arcs. Their general functionalities in a network are transport and coherence, respectively. Second, we introduce a betweenness centrality of arcs for each mode and see how the two modes are embedded in various real biological network data. We find that there is a trade-off relationship between the two centralities: if the value of one is large then the value of the other is small. This can be seen as a kind of division of labor in a network into transport on the network and coherence of the network. Finally, we propose an optimization model of networks based on a quality function involving intensities of the two modes in order to see how networks with the above trade-off relationship can emerge through evolution. We show that the trade-off relationship can be observed in the evolved networks only when the dynamic mode is dominant in the quality function by numerical simulations. We also show that the evolved networks have features qualitatively similar to real biological networks by standard complex network analysis.

Solving dynamical systems in neuromorphic hardware: simulation studies using balanced spiking networks

We aim at efficiently implementing and solving linear dynamical systems using neuromorphic hardware. For this task we used Deneve's balanced spiking network framework [1]. In this framework, recurrent spiking network of Leaky Integrate-and-Fire (LIF) neurons can track solution of a linear dynamical system by minimizing prediction error; weighted leaky integration is used to decode spike trains into a continuous signal. These networks have the following properties similar to properties of real biological networks: high trial-to-trial variability, asynchronous firing, tight balance between excitation and inhibition. Additionally, such networks could be implemented in silicon using analog neurons [2]. Analog neurons are power and cost-efficient, since they directly take advantage of physics laws. Noise induced during each analog operation is not cascaded but suppressed when a logical spike is emitted at the neuron's output. This hybrid scheme results in low area and power constraints for the neuron design while preserving computation efficiency. Creating neuromorphic microchips solving linear dynamical systems can greatly expand application scope of neuromorphic hardware, currently largely used for pattern recognition tasks. With some limitations of hardware implementations in mind, we set the following constraints for the network: 1) limited firing rate to keep energy cost low, 2) connection weights need to be set in advance (no learning), 3) the network should track the solution with a required precision, 4) network contains 1000-10000 of neurons. Our goal was to find out how to set the network parameters to achieve required solution accuracy. We considered several real life examples of dynamical systems. To learn how different network parameters affect accuracy of the solution, we ran Brian/Python simulations with different weight matrices, network noises, simulation time steps and leak constants. In this kind of networks, with each spike, the estimate of the solution is corrected roughly by the corresponding neuron's output weight vector. Testing showed that the network produces much more accurate results if the output weight vectors of all neurons have approximately equal euclidean norm. These two facts would cause problems with dynamical systems in which system variables exhibit different dynamics. The solution is to scale the variables of the dynamical systems using a linear transform chosen according to the following requirements: 1) the result of each spike should be large enough relative to the solution size so that the network can keep up with all variables' rate of change and decoder leaks under the condition of limited firing rate, 2) it should be small enough to avoid large relative error, 3) firing thresholds should be small enough compared to the input so that neuron leaks don't create a large error. We found that the network performs well with a reasonable amount of network noise, and, as expected, performance increases with the number of neurons. In our tests, normalized root-mean square error was under 0.01. These results have real-life applications, such as systems observation.

Graphlet-based measures are suitable for biological network comparison

Large amounts of biological network data exist for many species. Analogous to sequence comparison, network comparison aims to provide biological insight. Graphlet-based methods are proving to be useful in this respect. Recently some doubt has arisen concerning the applicability of graphlet-based measures to low edge density networks-in particular that the methods are 'unstable'-and further that no existing network model matches the structure found in real biological networks. We demonstrate that it is the model networks themselves that are 'unstable' at low edge density and that graphlet-based measures correctly reflect this instability. Furthermore, while model network topology is unstable at low edge density, biological network topology is stable. In particular, one must distinguish between average density and local density. While model networks of low average edge densities also have low local edge density, that is not the case with protein-protein interaction (PPI) networks: real PPI networks have low average edge density, but high local edge densities, and hence, they (and thus graphlet-based measures) are stable on these networks. Finally, we use a recently devised non-parametric statistical test to demonstrate that PPI networks of many species are well-fit by several models not previously tested. In addition, we model several viral PPI networks for the first time and demonstrate an exceptionally good fit between the data and theoretical models.

Rosen’s (M,R) system in process algebra

BackgroundRobert Rosen’s Metabolism-Replacement, or (M,R), system can be represented as a compact network structure with a single source and three products derived from that source in three consecutive reactions. (M,R) has been claimed to be non-reducible to its components and algorithmically non-computable, in the sense of not being evaluable as a function by a Turing machine. If (M,R)-like structures are present in real biological networks, this suggests that many biological networks will be non-computable, with implications for those branches of systems biology that rely on in silico modelling for predictive purposes.ResultsWe instantiate (M,R) using the process algebra Bio-PEPA, and discuss the extent to which our model represents a true realization of (M,R). We observe that under some starting conditions and parameter values, stable states can be achieved. Although formal demonstration of algorithmic computability remains elusive for (M,R), we discuss the extent to which our Bio-PEPA representation of (M,R) allows us to sidestep Rosen’s fundamental objections to computational systems biology.ConclusionsWe argue that the behaviour of (M,R) in Bio-PEPA shows life-like properties.

Efficient Routing Using Temporal Distance in Intermittently Connected Mobile Ad-hoc Networks

The analysis of real social, biological and technological networks has attracted a lot of attention as technological advances have given us a wealth of empirical data. For, analysis and investigation time varying graphs are used to understand the relationship, contact duration, repeated occurrence of contact. It is under exploring in intermittently connected networks. Now, by extending the same concept in intermittent networks, the efficiency of the routing protocol can be improved. This paper discusses about the temporal characterizing algorithm. Such characterization can help in accurately understanding dynamic behaviors and taking appropriate routing decisions. Therefore, the present research provokes exploring different possibilities of utilizing the same time varying network analyses and designing an Adaptive Routing protocol using temporal distance metric. The adaptive routing protocol is implemented using ONE simulator and is compared with the Epidemic and PropHET for delivery ratio, overhead and the number of dropped messages. The result reveals that Adaptive routing performs better than Epidemic and PropHET for real and synthetic datasets.

Boosted network classifiers for local feature selection.

Like all models, network feature selection models require that assumptions be made on the size and structure of the desired features. The most common assumption is sparsity, where only a small section of the entire network is thought to produce a specific phenomenon. The sparsity assumption is enforced through regularized models, such as the lasso. However, assuming sparsity may be inappropriate for many real-world networks, which possess highly correlated modules. In this paper, we illustrate two novel optimization strategies, namely, boosted expectation propagation (BEP) and boosted message passing (BMP), which directly use the network structure to estimate the parameters of a network classifier. BEP and BMP are ensemble methods that seek to optimize classification performance by combining individual models built upon local network features. Neither BEP nor BMP assumes a sparse solution, but instead they seek a weighted average of all network features where the weights are used to emphasize all features that are useful for classification. In this paper, we compare BEP and BMP with network-regularized logistic regression models on simulated and real biological networks. The results show that, where highly correlated network structure exists, assuming sparsity adversely effects the accuracy and feature selection power of the network classifier.

Parsimonious reconstruction of network evolution

BackgroundUnderstanding the evolution of biological networks can provide insight into how their modular structure arises and how they are affected by environmental changes. One approach to studying the evolution of these networks is to reconstruct plausible common ancestors of present-day networks, allowing us to analyze how the topological properties change over time and to posit mechanisms that drive the networks’ evolution. Further, putative ancestral networks can be used to help solve other difficult problems in computational biology, such as network alignment.ResultsWe introduce a combinatorial framework for encoding network histories, and we give a fast procedure that, given a set of gene duplication histories, in practice finds network histories with close to the minimum number of interaction gain or loss events to explain the observed present-day networks. In contrast to previous studies, our method does not require knowing the relative ordering of unrelated duplication events. Results on simulated histories and real biological networks both suggest that common ancestral networks can be accurately reconstructed using this parsimony approach. A software package implementing our method is available under the Apache 2.0 license at http://cbcb.umd.edu/kingsford-group/parana.ConclusionsOur parsimony-based approach to ancestral network reconstruction is both efficient and accurate. We show that considering a larger set of potential ancestral interactions by not assuming a relative ordering of unrelated duplication events can lead to improved ancestral network inference.

The architecture of dynamic reservoir in the echo state network

Echo state network (ESN) has recently attracted increasing interests because of its superior capability in modeling nonlinear dynamic systems. In the conventional echo state network model, its dynamic reservoir (DR) has a random and sparse topology, which is far from the real biological neural networks from both structural and functional perspectives. We hereby propose three novel types of echo state networks with new dynamic reservoir topologies based on complex network theory, i.e., with a small-world topology, a scale-free topology, and a mixture of small-world and scale-free topologies, respectively. We then analyze the relationship between the dynamic reservoir structure and its prediction capability. We utilize two commonly used time series to evaluate the prediction performance of the three proposed echo state networks and compare them to the conventional model. We also use independent and identically distributed time series to analyze the short-term memory and prediction precision of these echo state networks. Furthermore, we study the ratio of scale-free topology and the small-world topology in the mixed-topology network, and examine its influence on the performance of the echo state networks. Our simulation results show that the proposed echo state network models have better prediction capabilities, a wider spectral radius, but retain almost the same short-term memory capacity as compared to the conventional echo state network model. We also find that the smaller the ratio of the scale-free topology over the small-world topology, the better the memory capacities.

Uncovering missing links with cold ends

To evaluate the performance of prediction of missing links, the known data are randomly divided into two parts, the training set and the probe set. We argue that this straightforward and standard method may lead to terrible bias, since in real biological and information networks, missing links are more likely to be links connecting low-degree nodes. We therefore study how to uncover missing links with low-degree nodes, namely links in the probe set are of lower degree products than a random sampling. Experimental analysis on ten local similarity indices and four disparate real networks reveals a surprising result that the Leicht–Holme–Newman index [E.A. Leicht, P. Holme, M.E.J. Newman, Vertex similarity in networks, Phys. Rev. E 73 (2006) 026120] performs the best, although it was known to be one of the worst indices if the probe set is a random sampling of all links. We further propose an parameter-dependent index, which considerably improves the prediction accuracy. Finally, we show the relevance of the proposed index to three real sampling methods: acquaintance sampling, random-walk sampling and path-based sampling.

Topology and energy transport in networks of interacting photosynthetic complexes

We address the role of topology in the energy transport process that occurs in networks of photosynthetic complexes. We take inspiration from light-harvesting networks present in purple bacteria and simulate an incoherent dissipative energy transport process on more general and abstract networks, considering both regular structures (Cayley trees and hyperbranched fractals) and randomly generated ones. We focus on the the two primary light-harvesting complexes of purple bacteria, i.e., the LH1 and LH2, and we use network-theoretical centrality measures in order to select different LH1 arrangements. We show that different choices cause significant differences in the transport efficiencies, and that for regular networks, centrality measures allow us to identify arrangements that ensure transport efficiencies which are better than those obtained with a random disposition of the complexes. The optimal arrangements strongly depend on the dissipative nature of the dynamics and on the topological properties of the networks considered, and depending on the latter, they are achieved by using global versus local centrality measures. For randomly generated networks, a random arrangement of the complexes already provides efficient transport, and this suggests the process is strong with respect to limited amount of control in the structure design and to the disorder inherent in the construction of randomly assembled structures. Finally, we compare the networks considered with the real biological networks and find that the latter have in general better performances, due to their higher connectivity, but the former with optimal arrangements can mimic the real networks' behavior for a specific range of transport parameters. These results show that the use of network-theoretical concepts can be crucial for the characterization and design of efficient artificial energy transport networks.

A multilevel layout algorithm for visualizing physical and genetic interaction networks, with emphasis on their modular organization

BackgroundGraph drawing is an integral part of many systems biology studies, enabling visual exploration and mining of large-scale biological networks. While a number of layout algorithms are available in popular network analysis platforms, such as Cytoscape, it remains poorly understood how well their solutions reflect the underlying biological processes that give rise to the network connectivity structure. Moreover, visualizations obtained using conventional layout algorithms, such as those based on the force-directed drawing approach, may become uninformative when applied to larger networks with dense or clustered connectivity structure.MethodsWe implemented a modified layout plug-in, named Multilevel Layout, which applies the conventional layout algorithms within a multilevel optimization framework to better capture the hierarchical modularity of many biological networks. Using a wide variety of real life biological networks, we carried out a systematic evaluation of the method in comparison with other layout algorithms in Cytoscape.ResultsThe multilevel approach provided both biologically relevant and visually pleasant layout solutions in most network types, hence complementing the layout options available in Cytoscape. In particular, it could improve drawing of large-scale networks of yeast genetic interactions and human physical interactions. In more general terms, the biological evaluation framework developed here enables one to assess the layout solutions from any existing or future graph drawing algorithm as well as to optimize their performance for a given network type or structure.ConclusionsBy making use of the multilevel modular organization when visualizing biological networks, together with the biological evaluation of the layout solutions, one can generate convenient visualizations for many network biology applications.

Centrality prediction in dynamic human contact networks

Real technological, social and biological networks evolve over time. Predicting their future topology has applications to epidemiology, targeted marketing, network reliability and routing in ad hoc and peer-to-peer networks. The key problem for such applications is usually to identify the nodes that will be in more important positions in the future. Previous researchers had used ad hoc prediction functions. In this paper, we evaluate ways of predicting a node’s future importance under three important metrics, namely degree, closeness centrality, and betweenness centrality, using empirical data on human contact networks collected using mobile devices. We find that node importance is highly predictable due to both periodic and legacy effects of human social behaviour, and we design reasonable prediction functions. However human behaviour is not the same in all circumstances: the centrality of students at Cambridge is best correlated both daily and hourly, no doubt due to hourly lecture schedules, while academics at conferences exhibit rather flat closeness centrality, no doubt because conference attendees are generally trying to speak to new people at each break. This highlights the utility of having a number of different metrics for centrality in dynamic networks, so as to identify typical patterns and predict behaviour. We show that the best-performing prediction functions are 25% more accurate on average than simply using the previous centrality value. These prediction functions can be efficiently computed in linear time, and are thus practical for processing dynamic networks in real-time.

Degree-Based Sampling Method with Partition-Based Subgraph Finder for Larger Motif Detection

Network motifs are subnetworks that appear in the network far more frequently than in randomized networks. They have gathered much attention for uncovering structural design principles of complex networks. One of the previous approaches for motif detection is sampling method, in- troduced to perform the computational challenging task. However, it suffers from sampling bias and probability assignment. In addition, subgraph search, being very time-consuming, is a critical process in motif detection as we need to enumerate subgraphs of given sizes in the original input graph and an ensemble of random generated graphs. Therefore, we present a Degree-based Sampling Method with Partition-based Subgraph Finder for larger motif detection. Inspired by the intrinsic feature of real biological networks, Degree-based Sampling is a new solution for probability assignment based on degree. And, Partition-based Subgraph Finder takes its inspiration from the idea of partition, which improves computational efficiency and lowers space consumption. Experimental study on UETZ and E.COLI data set shows that the proposed method achieves more accuracy and efficiency than previous methods and scales better with increasing subgraph size.

METABOLIC NETWORKS ROBUSTNESS: THEORY, SIMULATIONS AND RESULTS

Metabolic networks are composed of several functional modules, reproducing metabolic pathways and describing the entire cellular metabolism of an organism. In the last decade, an enormous interest has grown for the study of tolerance to errors and attacks in metabolic networks. Studies on their robustness have suggested that metabolic networks are tolerant to errors, but very vulnerable to targeted attacks against highly connected nodes. However, many findings on metabolic networks suggest that the above classification is too simple and imprecise, since hub node attacks can be by-passed if alternative metabolic paths can be exploited. On the contrary, non-hub nodes attacks can affect cell survival when the node is the only path within a functional module. In this paper an integrated approach for metabolic networks robustness analysis is presented. With more details, statistical, topological, and functional analysis are used together to evaluate metabolic network behavior under normal operation conditions and under random or targeted attacks. Two real biological metabolic networks have been used to test the effectiveness of the proposed approach.

Global relative parameter sensitivities of the feed-forward loops in genetic networks

It is well known that the feed-forward loops (FFLs) are typical network motifs in many real world biological networks. The structures, functions, as well as noise characteristics of FFLs have received increasing attention over the last decade. This paper aims to further investigate the global relative parameter sensitivities (GRPS) of FFLs in genetic networks modeled by Hill kinetics by introducing a simple novel approach. Our results indicate that: (i) for the coherent FFLs (CFFLs), the most abundant type 1 configuration ( C1) is the most globally sensitive to system parameters, while for the incoherent FFLs (IFFLs), the most abundant type 1 configuration ( I1) is the least globally sensitive to system parameters; (ii) the less noisy of a FFL configuration, the more globally sensitive of this circuit to its parameters; and (iii) the most abundant FFL configurations are often either the least sensitive (robust) to system parameters variation (IFFLs) or the least noisy (CFFLs). Therefore, the above results can well explain the reason why FFLs are network motifs and are selected by nature in evolution. Furthermore, the proposed GRPS approach sheds some light on the potential real world applications, such as the synthetic genetic circuits, predicting the effect of interventions in medicine and biotechnology, and so on.

NetDS: a Cytoscape plugin to analyze the robustness of dynamics and feedforward/feedback loop structures of biological networks

NetDS is a novel Cytoscape plugin that conveniently simulates dynamics related to robustness, and examines structural properties with respect to feedforward/feedback loops. It can evaluate how robustly a network sustains a stable state against mutations by employing a Boolean network model. In addition, the plugin can examine all feedforward/feedback loops appearing in a network and determine whether or not a pair of loops is coupled. Random networks can also be generated to evaluate whether or not an interesting finding in real biological networks is significantly random. NetDS is freely available for non-commercial purposes at http://netds.sourceforge.net/. kwonyk@ulsan.ac.kr Supplementary data are available at Bioinformatics online.

Enrichment and aggregation of topological motifs are independent organizational principles of integrated interaction networks

Topological network motifs represent functional relationships within and between regulatory and protein-protein interaction networks. Enriched motifs often aggregate into self-contained units forming functional modules. Theoretical models for network evolution by duplication-divergence mechanisms and for network topology by hierarchical scale-free networks have suggested a one-to-one relation between network motif enrichment and aggregation, but this relation has never been tested quantitatively in real biological interaction networks. Here we introduce a novel method for assessing the statistical significance of network motif aggregation and for identifying clusters of overlapping network motifs. Using an integrated network of transcriptional, posttranslational and protein-protein interactions in yeast we show that network motif aggregation reflects a local modularity property which is independent of network motif enrichment. In particular our method identified novel functional network themes for a set of motifs which are not enriched yet aggregate significantly and challenges the conventional view that network motif enrichment is the most basic organizational principle of complex networks.

Fully parallel ANN-based arrhythmia classifier on a single-chip FPGA: FPAAC

Recognition of cardiac arrhythmias by electrocardiogram (ECG) is an important issue for diagnosis of cardiac abnormalities. Many studies on recognition of cardiac arrhythmias by ECG, using various techniques, have been performed in the past 20 years. Artificial neural networks (ANNs) are the most widely used tool in medical diagnosis systems (MDS) because of their powerful prediction characteristics. An ANN model is inspired by real biological neural networks, with a parallel structure that is potentially fast for computation. However, the suggested ANN architectures in the literature can only be run sequentially, on powerful processors, due to their complexity. Our approach enables the implementation of a simple ANN architecture with lower requirements for hardware resources. The features of the ECG signal are reduced dramatically using principle component analysis (PCA) while keeping the error rate of the ANN at an acceptable level, near 5%. To enable the implementation of real ANN models on parallel devices, the features of the ECG signal that are applied to the ANN inputs must be reduced. In this study, field programmable gate arrays (FPGA) implementation of a fully parallel, fault-tolerant ANN for ECG arrhythmia classification (FPAAC) is realized. An ANN model, which consists of 8 inputs, a hidden layer with 2 neurons, and 1 output neuron, is implemented on an FPGA using IEEE-754 32-bit floating-point numerical representation. FPAAC classifies 3 classes of arrhythmia, premature ventricular contraction (PVC), fusion (F), and normal (N) beats, and its accuracy is 97.66%. The ECG records used in this work were taken from the MIT-BIH arrhythmia database.

‘Non-destructive’ biocomputing security system based on gas-controlled biofuel cell and potentially used for intelligent medical diagnostics

Biofuel cells (BFCs) based on enzymes and microbes are the promising future alternative sources of sustainable electrical energy under mild conditions (i.e. ambient temperature and neutral pH). By combining the adaptive behavior of BFCs self-regulating energy release with the versatility of biocomputing, we construct a novel gas-controlled biocomputing security system, which could be used as the potential implantable self-powered and 'smart' medical system with the logic diagnosis aim. We have demonstrated a biocomputing security system based on BFCs. Due to the unique 'RESET' reagent of N(2) applied in this work, the prepared biocomputing security system can be reset and cycled for a large number of times with no 'RESET' reagent-based 'waste'. This would be advantageous for the potential practical applications of such keypad lock as well as the development of biocomputing security devices. In order to validate the universality of the system and also to harvest energy directly from biofuels with enhanced power output, we replace the glucose with orange juice as the biofuel to operate BFCs-based biocomputing system, which also possesses the function of keypad lock. In addition, by introducing BFCs into the biocomputing security system, the adaptive behavior of the BFCs self-regulating the power release would be an immense advantage of such security keypad lock devices in potential self-powered implantable medical systems. The designed sequence gives the maximum power output and discriminate itself from the rest of the sequences. From this, we find that maximizing the dimensionless ratio of gap versus SD of the power output spectrum (a funnel in power outputs) gives the quantitative optimal design criterion. Therefore, our construction here may also provide a practical example and microscopic structural basis for mimicking the real biological network systems and bridge the gaps between the theoretical concepts and experiments important for biomolecular systems and synthetic biology.

Local network-based measures to assess the inferability of different regulatory networks

The purpose of this study is to compare the inferability of various synthetic as well as real biological regulatory networks. In order to assess differences we apply local network-based measures. That means, instead of applying global measures, we investigate and assess an inference algorithm locally, on the level of individual edges and subnetworks. We demonstrate the behaviour of our local network-based measures with respect to different regulatory networks by conducting large-scale simulations. As inference algorithm we use exemplarily ARACNE. The results from our exploratory analysis allow us not only to gain new insights into the strength and weakness of an inference algorithm with respect to characteristics of different regulatory networks, but also to obtain information that could be used to design novel problem-specific statistical estimators.