Dissecting landscape art history with information theory.

Painting has played a major role in human expression, evolving subject to a complex interplay of representational conventions, social interactions, and a process of historization. From individual qualitative work of art historians emerges a metanarrative that remains difficult to evaluate in its validity regarding emergent macroscopic and underlying microscopic dynamics. The full scope of granular data, the summary statistics, and consequently, also their bias simply lie beyond the cognitive limit of individual qualitative human scholarship. Yet, a more quantitative understanding is still lacking, driven by a lack of data and a persistent dominance of qualitative scholarship in art history. Here, we show that quantitative analyses of creative processes in landscape painting can shed light, provide a systematic verification, and allow for questioning the emerging metanarrative. Using a quasicanonical benchmark dataset of 14,912 landscape paintings, covering a period from the Western renaissance to contemporary art, we systematically analyze the evolution of compositional proportion via a simple yet coherent information-theoretic dissection method that captures iterations of the dominant horizontal and vertical partition directions. Tracing frequency distributions of seemingly preferred compositions across several conceptual dimensions, we find that dominant dissection ratios can serve as a meaningful signature to capture the unique compositional characteristics and systematic evolution of individual artist bodies of work, creation date time spans, and conventional style periods, while concepts of artist nationality remain problematic. Network analyses of individual artists and style periods clarify their rhizomatic confusion while uncovering three distinguished yet nonintuitive supergroups that are meaningfully clustered in time.

Unexpected advantages of exploitation for target searches in complex networks.

Exploitation universally emerges in various decision-making contexts, e.g., animals foraging, web surfing, the evolution of scientists' research topics, and our daily lives. Despite its ubiquity, exploitation, which refers to the behavior of revisiting previous experiences, has often been considered to delay the search process of finding a target. In this paper, we investigate how exploitation affects search performance by applying a non-Markovian random walk model, where a walker randomly revisits a previously visited node using long-term memory. We analytically study two broad forms of network structures, namely, (i) clique-like networks and (ii) lollipop-like networks and find that exploitation can significantly improve search performance in lollipop-like networks, whereas it hinders target search in clique-like networks. Moreover, we numerically verify that exploitation can reduce the time needed to fully explore the underlying networks using 550 diverse real-world networks. Based on the analytic result, we define the lollipop-likeness of a network and observe a positive relationship between the advantage of exploitation and lollipop-likeness.

Encoding Multiple Virtual Signals in DNA Barcodes with Single-Molecule FRET.

DNA barcoding provides a way to label a myriad of different biological molecules using the extreme programmability in DNA sequence synthesis. Fluorescence imaging is presumably the most easy-to-access method for DNA barcoding, yet large spectral overlaps between fluorescence dyes severely limit the numbers of barcodes that can be detected simultaneously. We here demonstrate the use of single-molecule fluorescence resonance energy transfer (FRET) to encode virtual signals in DNA barcodes using conventional two-color fluorescence microscopy. By optimizing imaging and biochemistry conditions for weak DNA hybridization events, we markedly enhanced accuracy in our determination of the single-molecule FRET efficiency exhibited by each binding event between DNA barcode sequences. This allowed us to unambiguously differentiate six DNA barcodes encoding different FRET values without involving any probe sequence exchanges. Our method can be directly incorporated with previous DNA barcode techniques, and may thus be widely adopted to expand the signal space of DNA barcoding.

Learning Entropy Production via Neural Networks.

This Letter presents a neural estimator for entropy production (NEEP), that estimates entropy production (EP) from trajectories of relevant variables without detailed information on the system dynamics. For steady state, we rigorously prove that the estimator, which can be built up from different choices of deep neural networks, provides stochastic EP by optimizing the objective function proposed here. We verify the NEEP with the stochastic processes of the bead spring and discrete flashing ratchet models and also demonstrate that our method is applicable to high-dimensional data and can provide coarse-grained EP for Markov systems with unobservable states.

Chromosome-like organization of an asymmetrical ring polymer confined in a cylindrical space.

To what extent does a confined polymer show chromosome-like organization? Using molecular dynamics simulations, we study a model Escherichia coli (E. coli) chromosome: an asymmetrical ring polymer, formed by small monomers on one side and big monomers on the other confined in a concentric-shell or simple cylinder with closed ends. The ring polymer is organized in the way observed for the E. coli chromosome: if the big monomers are assumed to be localized in the inner cylinder, the two "subchains" forming the ring are spontaneously partitioned in a parallel orientation with the "body" (big-monomer) chain linearly organized with a desired precision and the crossing (small-monomer) chain residing preferentially in the peripheral region. Furthermore, we show that the introduction of a "fluctuating boundary" between the two subchains leads to a double-peak distribution of ter-proximate loci, as seen in experiments, which would otherwise remain single-peaked. In a simple cylinder, however, a chromosome-like organization of the ring polymer typically requires an external mechanism such as cell-wall attachment. Finally, our results clarify to what degree the spatial organization of the chromosomes can be accomplished solely by ring asymmetry and anisotropic confinement.

A polymer in a crowded and confined space: effects of crowder size and poly-dispersity.

DNA compaction in a bacterial cell is in part carried out by entropic (depletion) forces induced by "free" proteins or crowding particles in the cytoplasm. Indeed, recent in vitro experiments highlight these effects by showing that they alone can condense the E. coli chromosome to its in vivo size. Using molecular dynamics simulations and a theoretical approach, we study how a flexible chain molecule can be compacted by crowding particles with variable sizes in a (cell-like) cylindrical space. Our results show that with smaller crowding agents the compaction occurs at a lower volume fraction but at a larger concentration such that doubling their size is equivalent to increasing their concentration fourfold. Similarly, the effect of polydispersity can be correctly mimicked by adjusting the size of crowders in a homogeneous system. Under different conditions, however, crowding particles can induce chain adsorption onto the cylinder wall, stretching the chain, which would otherwise remain condensed.

Hidden multiscale organization and robustness of real multiplex networks.

Hidden geometry enables the investigation of complex networks at different scales. Extending this framework to multiplex networks, we uncover a different kind of mesoscopic organization in real multiplex systems, named clan, a group of nodes that preserve local geometric arrangements across layers. Furthermore, we reveal the intimate relationship between the unfolding of clan structure and mutual percolation against targeted attacks, leading to an ambivalent role of clans: making a system fragile yet less prone to complete shattering. Finally, we confirm the correlation between the multiscale nature of geometric organization and the overall robustness. Our findings expand the significance of hidden geometry in network function, while also highlighting potential pitfalls in evaluating and controlling catastrophic failure of multiplex systems.

Spontaneous emergence of rudimentary music detectors in deep neural networks.

Music exists in almost every society, has universal acoustic features, and is processed by distinct neural circuits in humans even with no experience of musical training. However, it remains unclear how these innate characteristics emerge and what functions they serve. Here, using an artificial deep neural network that models the auditory information processing of the brain, we show that units tuned to music can spontaneously emerge by learning natural sound detection, even without learning music. The music-selective units encoded the temporal structure of music in multiple timescales, following the population-level response characteristics observed in the brain. We found that the process of generalization is critical for the emergence of music-selectivity and that music-selectivity can work as a functional basis for the generalization of natural sound, thereby elucidating its origin. These findings suggest that evolutionary adaptation to process natural sounds can provide an initial blueprint for our sense of music.

Fundamental structural constraint of random scale-free networks.

We study the structural constraint of random scale-free networks that determines possible combinations of the degree exponent γ and the upper cutoff k(c) in the thermodynamic limit. We employ the framework of graphicality transitions proposed by Del Genio and co-workers [Phys. Rev. Lett. 107, 178701 (2011)], while making it more rigorous and applicable to general values of k(c). Using the graphicality criterion, we show that the upper cutoff must be lower than k(c)∼N(1/γ) for γ<2, whereas any upper cutoff is allowed for γ>2. This result is also numerically verified by both the random and deterministic sampling of degree sequences.

Scaling laws between population and facility densities.

When a new facility like a grocery store, a school, or a fire station is planned, its location should ideally be determined by the necessities of people who live nearby. Empirically, it has been found that there exists a positive correlation between facility and population densities. In the present work, we investigate the ideal relation between the population and the facility densities within the framework of an economic mechanism governing microdynamics. In previous studies based on the global optimization of facility positions in minimizing the overall travel distance between people and facilities, it was shown that the density of facility D and that of population rho should follow a simple power law D approximately rho(2/3). In our empirical analysis, on the other hand, the power-law exponent alpha in D approximately rho(alpha) is not a fixed value but spreads in a broad range depending on facility types. To explain this discrepancy in alpha, we propose a model based on economic mechanisms that mimic the competitive balance between the profit of the facilities and the social opportunity cost for populations. Through our simple, microscopically driven model, we show that commercial facilities driven by the profit of the facilities have alpha = 1, whereas public facilities driven by the social opportunity cost have alpha = 2/3. We simulate this model to find the optimal positions of facilities on a real U.S. map and show that the results are consistent with the empirical data.

Unraveling hidden interactions in complex systems with deep learning.

Rich phenomena from complex systems have long intrigued researchers, and yet modeling system micro-dynamics and inferring the forms of interaction remain challenging for conventional data-driven approaches, being generally established by scientists with human ingenuity. In this study, we propose AgentNet, a model-free data-driven framework consisting of deep neural networks to reveal and analyze the hidden interactions in complex systems from observed data alone. AgentNet utilizes a graph attention network with novel variable-wise attention to model the interaction between individual agents, and employs various encoders and decoders that can be selectively applied to any desired system. Our model successfully captured a wide variety of simulated complex systems, namely cellular automata (discrete), the Vicsek model (continuous), and active Ornstein-Uhlenbeck particles (non-Markovian) in which, notably, AgentNet's visualized attention values coincided with the true variable-wise interaction strengths and exhibited collective behavior that was absent in the training data. A demonstration with empirical data from a flock of birds showed that AgentNet could identify hidden interaction ranges exhibited by real birds, which cannot be detected by conventional velocity correlation analysis. We expect our framework to open a novel path to investigating complex systems and to provide insight into general process-driven modeling.

Inertial effects on the Brownian gyrator.

The recent interest into the Brownian gyrator has been confined chiefly to the analysis of Brownian dynamics both in theory and experiment despite the applicability of general cases with definite mass. Considering mass explicitly in the solution of the Fokker-Planck equation and Langevin dynamics simulations, we investigate how inertia can change the dynamics and energetics of the Brownian gyrator. In the Langevin model, the inertia reduces the nonequilibrium effects by diminishing the declination of the probability density function and the mean of a specific angular momentum, j_{θ}, as a measure of rotation. Another unique feature of the Langevin description is that rotation is maximized at a particular anisotropy while the stability of the rotation is minimized at a particular anisotropy or mass. Our results suggest that the Langevin dynamics description of the Brownian gyrator is intrinsically different from that with Brownian dynamics. In addition, j_{θ} is proven to be essential and convenient for estimating stochastic energetics such as heat currents and entropy production even in the underdamped regime.

Impact of environmental changes on the dynamics of temporal networks.

Dynamics of complex social systems has often been described in the framework of temporal networks, where links are considered to exist only at the moment of interaction between nodes. Such interaction patterns are not only driven by internal interaction mechanisms, but also affected by environmental changes. To investigate the impact of the environmental changes on the dynamics of temporal networks, we analyze several face-to-face interaction datasets using the multiscale entropy (MSE) method to find that the observed temporal correlations can be categorized according to the environmental similarity of datasets such as classes and break times in schools. By devising and studying a temporal network model considering a periodically changing environment as well as a preferential activation mechanism, we numerically show that our model could successfully reproduce various empirical results by the MSE method in terms of multiscale temporal correlations. Our results demonstrate that the environmental changes can play an important role in shaping the dynamics of temporal networks when the interactions between nodes are influenced by the environment of the systems.

Quantumness and thermodynamic uncertainty relation of the finite-time Otto cycle.

To reveal the role of the quantumness in the Otto cycle and to discuss the validity of the thermodynamic uncertainty relation (TUR) in the cycle, we study the quantum Otto cycle and its classical counterpart. In particular, we calculate exactly the mean values and relative error of thermodynamic quantities. In the quasistatic limit, quantumness reduces the productivity and precision of the Otto cycle compared to that in the absence of quantumness, whereas in the finite-time mode, it can increase the cycle's productivity and precision. Interestingly, as the strength (heat conductance) between the system and the bath increases, the precision of the quantum Otto cycle overtakes that of the classical one. Testing the conventional TUR of the Otto cycle, in the region where the entropy production is large enough, we find a tighter bound than that of the conventional TUR. However, in the finite-time mode, both quantum and classical Otto cycles violate the conventional TUR in the region where the entropy production is small. This implies that another modified TUR is required to cover the finite-time Otto cycle. Finally, we discuss the possible origin of this violation in terms of the uncertainty products of the thermodynamic quantities and the relative error near resonance conditions.

Finite-time quantum Otto engine: Surpassing the quasistatic efficiency due to friction.

In finite-time quantum heat engines, some work is consumed to drive a working fluid accompanying coherence, which is called "friction." To understand the role of friction in quantum thermodynamics, we present a couple of finite-time quantum Otto cycles with two different baths: Agarwal versus Lindbladian. We solve them exactly and compare the performance of the Agarwal engine with that of the Lindbladian engine. In particular, we find remarkable and counterintuitive results that the performance of the Agarwal engine due to friction can be much higher than that in the quasistatic limit with the Otto efficiency, and the power of the Lindbladian engine can be nonzero in the short-time limit. Based on additional numerical calculations of these outcomes, we discuss possible origins of such differences between two engines and reveal them. Our results imply that, even with an equilibrium bath, a nonequilibrium working fluid brings on the higher performance than what an equilibrium working fluid does.

Dynamics and directionality in complex networks.

We investigate how we can improve the synchronizability of complex networks simply by changing the link direction while conserving the local link weights and topology. Performing the linear stability analysis of synchronization and numerical simulation of the Kuramoto model in the directed networks, we find that while a random assignment of link directions generally weakens the degree of synchronization, a properly organized directionality can systematically enhance the network synchronization. In this respect, we suggest a simple method of changing the link direction according to the larger residual degree starting from small residual degree nodes. This result provides plausible applications to control the synchronizability of systems in various fields.

Early onset of structural inequality in the formation of collaborative knowledge in all Wikimedia projects.

The Wikimedia project, including Wikipedia, is one of the largest communal data sets and has served as a representative medium to convey collective knowledge in the twenty-first century. Researchers have believed that the analysis of these collaborative digital data sets provides a unique window into the processes of collaborative knowledge formation; yet, in reality, most previous studies have usually focused on its narrow subsets. Here, by analysing all 863 Wikimedia projects (various types and in different languages), we find evidence for a universal growth pattern in communal data formation. We observe that inequality arises early in the development of Wikimedia projects and stabilizes at high levels. To understand the mechanism behind the observed structural inequality, we develop an agent-based model that considers the characteristics of the editors and successfully reproduces the empirical results. Our findings from the Wikimedia projects data, along with other types of collaboration data, such as patents and academic papers, show that a small number of editors have a disproportionately large influence on the formation of collective knowledge. This analysis offers insights into how various collaboration environments can be sustained in the future.

Role of hubs in the synergistic spread of behavior.

The spread of behavior in a society has two major features: the synergy of multiple spreaders and the dominance of hubs. While strong synergy is known to induce mixed-order transitions (MOTs) at percolation, the effects of hubs on the phenomena are yet to be clarified. By analytically solving the generalized epidemic process on random scale-free networks with the power-law degree distribution p_{k}∼k^{-α}, we clarify how the dominance of hubs in social networks affects the conditions for MOTs. Our results show that, for α<4, an abundance of hubs drive MOTs, even if a synergistic spreading event requires an arbitrarily large number of adjacent spreaders. In particular, for 2<α<3, we find that a global cascade is possible even when only synergistic spreading events are allowed. These transition properties are substantially different from those of cooperative contagions, which are another class of synergistic cascading processes exhibiting MOTs.

Watching helical membrane proteins fold reveals a common N-to-C-terminal folding pathway.

To understand membrane protein biogenesis, we need to explore folding within a bilayer context. Here, we describe a single-molecule force microscopy technique that monitors the folding of helical membrane proteins in vesicle and bicelle environments. After completely unfolding the protein at high force, we lower the force to initiate folding while transmembrane helices are aligned in a zigzag manner within the bilayer, thereby imposing minimal constraints on folding. We used the approach to characterize the folding pathways of the Escherichia coli rhomboid protease GlpG and the human ß2-adrenergic receptor. Despite their evolutionary distance, both proteins fold in a strict N-to-C-terminal fashion, accruing structures in units of helical hairpins. These common features suggest that integral helical membrane proteins have evolved to maximize their fitness with cotranslational folding.

A protein interaction network associated with asthma.

Identifying candidate genes related to complex diseases or traits and mapping their relationships require a system-level analysis at a cellular scale. The objective of the present study is to systematically analyze the complex effects of interrelated genes and provide a framework for revealing their relationships in association with a specific disease (asthma in this case). We observed that protein-protein interaction (PPI) networks associated with asthma have a power-law connectivity distribution as many other biological networks have. The hub nodes and skeleton substructure of the result network are consistent with the prior knowledge about asthma pathways, and also suggest unknown candidate target genes associated with asthma, including GNB2L1, BRCA1, CBL, and VAV1. In particular, GNB2L1 appears to play a very important role in the asthma network through frequent interactions with key proteins in cellular signaling. This network-based approach represents an alternative method for analyzing the complex effects of candidate genes associated with complex diseases and suggesting a list of gene drug targets. The full list of genes and the analysis details are available in the following online supplementary materials: http://biosoft.kaist.ac.kr:8080/resources/asthma_ppi.