Complexity Thoughts: Issue #24
Unraveling complexity: building knowledge, one paper at a time
If you find Complexity Thoughts, click on the Like button, leave a comment, repost on Substack or share this post. It is the only feedback I can have for this free service.
The frequency and quality of this newsletter relies on social interactions. Thank you!
Dall-e 3 representation of this issue’s content
From the lab
Robustness and resilience of complex networks
A variety of biological, social and engineering complex systems can be defined in terms of units that exchange information through interaction networks, exhibiting diverse structural patterns such as heterogeneity, modularity and hierarchy. Owing to their interconnected nature, complex networks can amplify minor disruptions to a system-wide level, making it essential to understand their robustness against both external perturbations and internal failures. The study of complex networks’ robustness and resilience involves investigating phase transitions that usually depend on features such as degree connectivity, spatial embedding, interdependence and coupled dynamics. Network science offers a wide range of theoretical and computational methods for quantifying system robustness against perturbations, as well as grounded approaches to design robustness, identify early-warning signals and devise adaptive responses. These methods find application across a multitude of disciplines, including systems biology, systems neuroscience, engineering, and social and behavioural sciences.
Complex networks are ubiquitous: a cell, the human brain, a group of people and the Internet are all examples of interconnected many-body systems characterized by macroscopic properties that cannot be trivially deduced from those of their microscopic constituents. Such systems are exposed to both internal, localized, failures and external disturbances or perturbations. Owing to their interconnected structure, complex systems might be severely degraded, to the point of disintegration or systemic dysfunction. Examples include cascading failures, triggered by an initially localized overload in power systems, and the critical slowing downs of ecosystems which can be driven towards extinction. In recent years, this general phenomenon has been investigated by framing localized and systemic failures in terms of perturbations that can alter the function of a system. We capitalize on this mathematical framework to review theoretical and computational approaches to characterize robustness and resilience of complex networks. We discuss recent approaches to mitigate the impact of perturbations in terms of designing robustness, identifying early-warning signals and adapting responses. In terms of applications, we compare the performance of the state-of-the-art dismantling techniques, highlighting their optimal range of applicability for practical problems, and provide a repository with ready-to-use scripts, a much-needed tool set.
Neuroscience
Circular and unified analysis in network neuroscience
It’s simply impossible not to appreciate this short review piece by Mika Rubinov.
You do not know anything until you know why you know it.
— Clovis Andersen, The Principles of Private Detection (McCall Smith, 2007), cited in Sokal, 2010
Genuinely new discovery transcends existing knowledge. Despite this, many analyses in systems neuroscience neglect to test new speculative hypotheses against benchmark empirical facts. Some of these analyses inadvertently use circular reasoning to present existing knowledge as new discovery. Here, I discuss that this problem can confound key results and estimate that it has affected more than three thousand studies in network neuroscience over the last decade. I suggest that future studies can reduce this problem by limiting the use of speculative evidence, integrating existing knowledge into benchmark models, and rigorously testing proposed discoveries against these models. I conclude with a summary of practical challenges and recommendations.
Does the brain behave like a (complex) network? I. Dynamics
This is a well needed review!
It is therefore essential to evaluate the extent to which a network representation, particularly at system-level scales, can reveal fundamental aspects of brain dynamics, whether it can produce specific brain phenomenology, and ultimately whether it genuinely documents the way the brain carries out the functions it is supposed to fulfil, the extent to which they show robustness with respect to the way they are equipped with a network structure or the way in which the structure is parsed when analysing it.
Graph theory is now becoming a standard tool in system-level neuroscience. However, endowing observed brain anatomy and dynamics with a complex network structure does not entail that the brain actually works as a network. Asking whether the brain behaves as a network means asking whether network properties count. From the viewpoint of neurophysiology and, possibly, of brain physics, the most substantial issues a network structure may be instrumental in addressing relate to the influence of network properties on brain dynamics and to whether these properties ultimately explain some aspects of brain function. Here, we address the dynamical implications of complex network, examining which aspects and scales of brain activity may be understood to genuinely behave as a network. To do so, we first define the meaning of networkness, and analyse some of its implications. We then examine ways in which brain anatomy and dynamics can be endowed with a network structure and discuss possible ways in which network structure may be shown to represent a genuine organisational principle of brain activity, rather than just a convenient description of its anatomy and dynamics.
An energy costly architecture of neuromodulators for human brain evolution and cognition
Brain structures across a variety of species have evolved under similar organizational principles: nevertheless, they display remarkable diversity. Some species surpass humans in brain size (eg, Indian elephant), brain-to-body mass ratio (eg, mouse), or neuron count (e.g., the long-finned pilot whale). For some details and references, you could read the Is a large scale enough to guarantee emergent phenomena? section at the end of one of my past essays about emergent features in AI.
It is plausible to think that the aforementioned diversity implies that factors beyond mere structural scaling should contribute to the unique cognitive abilities exhibited by humans. One possibility is related to the metabolic profile of the brain connectome. In fact, our brain is quite a high-energy-demand organ, requiring a steady influx of energy. Remarkably, despite the high energy needs of the human brain relative to total body metabolism, there is a basic neuronal design that remains evolutionarily conserved across mammals, with a comparable signaling cost per neuron. Since for a primate of its size the human brain consists of the expected neuron and non-neuron counts, exhibiting a neuron distribution in the cerebral cortex akin to other species, where is the difference? The authors propose that variations in regional brain energy demands are influenced by the intensity of signaling within the connectome, rather than just the structural attributes of the brain.
Our findings suggest that the evolution of human cognition may have emerged not only from an overall larger brain, but particularly by the development of slow-acting neuromodulator circuits
In comparison to other species, the human brain exhibits one of the highest energy demands relative to body metabolism. It remains unclear whether this heightened energy demand uniformly supports an enlarged brain or if specific signaling mechanisms necessitate greater energy. We hypothesized that the regional distribution of energy demands will reveal signaling strategies that have contributed to human cognitive development. We measured the energy distribution within the brain functional connectome using multimodal brain imaging and found that signaling pathways in evolutionarily expanded regions have up to 67% higher energetic costs than those in sensory-motor regions. Additionally, histology, transcriptomic data, and molecular imaging independently reveal an up-regulation of signaling at G-protein-coupled receptors in energy-demanding regions. Our findings indicate that neuromodulator activity is predominantly involved in cognitive functions, such as reading or memory processing. This study suggests that an up-regulation of neuromodulator activity, alongside increased brain size, is a crucial aspect of human brain evolution.
Population dynamics and epidemics
Epidemic graph diagrams as analytics for epidemic control in the data-rich era
COVID-19 highlighted modeling as a cornerstone of pandemic response. But it also revealed that current models may not fully exploit the high-resolution data on disease progression, epidemic surveillance and host behavior, now available. Take the epidemic threshold, which quantifies the spreading risk throughout epidemic emergence, mitigation, and control. Its use requires oversimplifying either disease or host contact dynamics. We introduce the epidemic graph diagrams to overcome this by computing the epidemic threshold directly from arbitrarily complex data on contacts, disease and interventions. A grammar of diagram operations allows to decompose, compare, simplify models with computational efficiency, extracting theoretical understanding. We use the diagrams to explain the emergence of resistant influenza variants in the 2007–2008 season, and demonstrate that neglecting non-infectious prodromic stages of sexually transmitted infections biases the predicted epidemic risk, compromising control. The diagrams are general, and improve our capacity to respond to present and future public health challenges.
Evolution
Contingency, repeatability, and predictability in the evolution of a prokaryotic pangenome
The prokaryotic pangenome embodies the complete gene set of a prokaryotic species, integrating the core genome (shared by all strains), the accessory genome (present in some strains), and the unique genome (specific to individual strains). This concept is instrumental for understanding the diversity observed in prokaryotes and evolution. In a nutshell:
Prokaryotic genomes can dynamically gain and lose genes through horizontal gene transfer, gene duplication, and gene loss (all processes that foster genetic diversity)
The pangenome allows prokaryotes to adapt to very diverse environments, with accessory and unique genes that can confer specialized traits useful for adaptation
The structure of the pangenome is shaped by both environmental and evolutionary pressures, influencing microbial population dynamics and speciation.
Why is it interesting? Because it can offer novel insights into how microbial communities adapt and evolve, as well as on their ecological roles.
We propose that at least part of the pangenome can be understood as a set of genes with relationships that govern their likely cohabitants, analogous to an ecosystem’s set of interacting organisms
Different strains of the same prokaryotic species often show significant variation in gene content. Whether this variation is due to genetic drift or selection is not well understood. If the latter, we expect sets of genes to be consistently and repeatedly gained or lost together, or sequentially. We used machine learning to predict the presence of variable genes in a large set of Escherichia coli strains, using other variable genes as predictors. We find a large proportion of genes are predictable, suggesting selection plays a role in their acquisition, loss, and maintenance. We show that some genes are consistently associated with the presence or absence of others. These results have implications for understanding evolutionary dynamics in prokaryotic genomes.
Mutation enhances cooperation in direct reciprocity
Cooperation as a perfectly valid evolutionary force. Another interesting piece of work adding to our knowledge of evolution.
In our paper, cooperation does not evolve because individual strategies are sufficiently sophisticated. Instead, cooperation evolves because the evolutionary process is sufficiently erratic
Evolution is not only fierce competition but also cooperation. The latter can be seen as the main architect of biological complexity. Cooperation means that one individual pays a cost for another individual to receive a benefit. Cooperation is opposed by natural selection unless a mechanism for evolution of cooperation is at work. One such mechanism is direct reciprocity which utilizes repeated interactions between the same individuals. Here, we report that intermediate mutation rates, which lead to diverse communities of cooperators, dramatically enhance the levels of cooperation in competitive settings. Therefore, mutation “cooperates” with cooperation to maximize its beneficial effect on life.
Computational social science
Unsupervised embedding of trajectories captures the latent structure of scientific migration
Scientific migration is a fact for many researchers and scientists. By using a large-scale database of scientists’ trajectories inferred from publication records, the authors propose an interesting way to study the data by means of word2vec:
In revealing the correspondence between neural embeddings and the gravity model, the study of human migration can move beyond geographic and network-based models of migration, and instead leverage the high-order structure directly from individuals’ migration trajectories using these robust and efficient methods
We show that the word embedding technique word2vec is mathematically equivalent to the gravity law of mobility, making it ideal for learning dense representations from migration data that can be distributed, re-used, and studied. By treating locations analogously to words and trajectories to sentences, word2vec embeds each location into a vector space, where the distance reflects rates of migration according to the gravity law. We demonstrate the power of word2vec by applying it to the case of scientists’ migrations, for which it encodes information about culture, geography, and prestige at multiple layers of granularity. Our results lay a theoretical and methodological foundation for the application of neural embeddings to the study of migration.
Considering the current political climate in the United States, greater transparency in social recommendation systems and also auditing these systems for filter bubbles and rabbit holes of radicalization is timely and needed
YouTube’s algorithm is often accused of putting users in filter bubbles and generating rabbit holes of radicalization. However, evidence on these issues is inconclusive. We conduct a systematic audit of the platform using 100,000 sock puppets that allow us to isolate the influence of the algorithm in recommendations to ideologically congenial and increasingly extreme and problematic videos. YouTube’s algorithm recommends ideologically congenial content to partisan users, and congenial recommendations increase deeper in the recommendation trail for right-leaning users. Although we do not find meaningful increases in ideological extremity of recommendations, we show that a growing proportion of recommendations deeper in the recommendation trail come from extremist, conspiratorial, and otherwise problematic channels. This increase is most pronounced among the right-leaning users.
Tools
PyRates—A code-generation tool for modeling dynamical systems in biology and beyond: a Python-based software for modeling and analyzing differential equation systems via numerical methods. PyRates is specifically designed to account for the inherent complexity of biological systems.
Catalyst: Fast and flexible modeling of reaction networks: a flexible and feature-filled Julia library for modeling and high-performance simulation of chemical reaction networks (CRNs). Catalyst supports simulating stochastic chemical kinetics (jump process), chemical Langevin equation (stochastic differential equation), and reaction rate equation (ordinary differential equation) representations for CRNs.