Francesco Bullo; Pietro Hiram Guzzi

Recent years have seen a remarkable rise in the number of applications of network science in
the fields of biology and medicine. Networks in biology and medicine aim to represent the
organisation of living systems as a set of interacting elements at different scales, from the
subcellular to the population level. The initial application of network science in medicine
primarily focused on understanding the structure of protein interaction networks and using
these relationships to hypothesize new disease genes or novel therapeutic targets. At the same
time, network science has been widely applied in the context of molecular biology, for example
to model biological processes as gene regulatory networks. Now, with the continual influx
of biomedical big data, which is providing increasingly detailed information about various
aspects of molecular biology and medicine, the scale and scope of the network models used in
biology and medicine have skyrocketed. For example, improvements in medical imaging has
greatly facilitated the study of brain interaction networks.
Moving forward, it is imperative to develop approaches that holistically model the
complexity inherent in biological systems. Network science, in particular, has the potential to
answer critical questions in medicine that cannot be addressed through standard approaches.
By capitalizing on tools designed to quantify the fundamental properties of large-scale complex
systems, network science offers a complementary view to that of systems biology, which
tends to focus on basic mechanisms and small to medium-scale biological models. We believe
there has never been a better opportunity to employ network science to make sense of large,
complex biological systems and tackle some of medicine’s most challenging open questions. In this white paper, we review several key areas where network science has the best opportunity to contribute to medical applications and posit several critical future directions for the field.

Read the full article at: netscisociety.net

Matthew Brouillet, Georgi Yordanov Georgiev

Processes 2024, 12(12), 2937

Self-organization in complex systems is a process associated with reduced internal entropy and the emergence of structures that may enable the system to function more effectively and robustly in its environment and in a more competitive way with other states of the system or with other systems. This phenomenon typically occurs in the presence of energy gradients, facilitating energy transfer and entropy production. As a dynamic process, self-organization is best studied using dynamic measures and principles. The principles of minimizing unit action, entropy, and information while maximizing their total values are proposed as some of the dynamic variational principles guiding self-organization. The least action principle (LAP) is the proposed driver for self-organization; however, it cannot operate in isolation; it requires the mechanism of feedback loops with the rest of the system’s characteristics to drive the process. Average action efficiency (AAE) is introduced as a potential quantitative measure of self-organization, reflecting the system’s efficiency as the ratio of events to total action per unit of time. Positive feedback loops link AAE to other system characteristics, potentially explaining power–law relationships, quantity–AAE transitions, and exponential growth patterns observed in complex systems. To explore this framework, we apply it to agent-based simulations of ants navigating between two locations on a 2D grid. The principles align with observed self-organization dynamics, and the results and comparisons with real-world data appear to support the model. By analyzing AAE, this study seeks to address fundamental questions about the nature of self-organization and system organization, such as “Why and how do complex systems self-organize? What is organization and how organized is a system?”. We present AAE for the discussed simulation and whenever no external forces act on the system. Given so many specific cases in nature, the method will need to be adapted to reflect their specific interactions. These findings suggest that the proposed models offer a useful perspective for understanding and potentially improving the design of complex systems.

Read the full article at: www.mdpi.com

Edmonds, B., Hofstede, G. J., Koch, J., le Page, C., Lim, T., Lippe, M., Nöldeke, B., & van Delden, H.

Socio-Environmental Systems Modelling, 6, 18593.

Socio-Ecological System modelling projects are becoming increasingly complicated, with multiple actors and aspects being the norm. Such projects can cause problems for the modellers when this involves different elements, goals, philosophies, etc., all pulling in different directions – we call this “Chimaera Modelling.” Although such situations are common when you talk to modellers, they do not seem to be explicitly discussed in the literature. In this paper, we attempt to turn this perceived “inside” phenomenon into an “outside” phenomenon and to start a debate to increase transparency among the modelling community. We discuss the different aspects which may be relevant to this problem to start this debate, including: the underlying philosophy, modelling goals, extent of choice the modellers have, different stages of modelling, and kinds of actors that are involved. We further map out some of the dimensions with which Chimaera Modelling connects. We briefly discuss these and propose to the community as a whole to work on their methodological development, feasibility, risks and applicability as their resolution is far beyond the scope of this paper. We end with a brief description of the broad possible approaches to such situations. Our main message is a call for recognition of Chimaera Modelling as a likely side-effect of multi-stakeholder, multi-purpose projects, and to take this into account proactively at the project team level and be transparent about the tensions and contradictions that underly such modelling.

Read the full article at: sesmo.org

AKARSH KUMAR, CHRIS LU, LOUIS KIRSCH, YUJIN TANG, KENNETH STANLEY, PHILLIP ISOLA, DAVID HA

With the recent Nobel Prize awarded for radical advances in protein discovery, foundation models (FMs) for exploring large combinatorial spaces promise to revolutionize many scientific fields. Artificial Life (ALife) has not yet integrated FMs, thus presenting a major opportunity for the field to alleviate the historical burden of relying chiefly on manual design and trial-and-error to discover the configurations of lifelike simulations. This paper presents, for the first time, a successful realization of this opportunity using vision-language FMs. The proposed approach, called Automated Search for Artificial Life (ASAL), (1) finds simulations that produce target phenomena, (2) discovers simulations that generate temporally open-ended novelty, and (3) illuminates an entire space of interestingly diverse simulations. Because of the generality of FMs, ASAL works effectively across a diverse range of ALife substrates including Boids, Particle Life, Game of Life, Lenia, and Neural Cellular Automata. A major result highlighting the potential of this technique is the discovery of previously unseen Lenia and Boids lifeforms, as well as cellular automata that are open-ended like Conway’s Game of Life. Additionally, the use of FMs allows for the quantification of previously qualitative phenomena in a human-aligned way. This new paradigm promises to accelerate ALife research beyond what is possible through human ingenuity alone.

Read the full article at:sakana.ai

Dino Pedreschi, Luca Pappalardo, Emanuele Ferragina, Ricardo Baeza-Yates, Albert-László Barabási, Frank Dignum, Virginia Dignum, Tina Eliassi-Rad, Fosca Giannotti, János Kertész, Alistair Knott, Yannis Ioannidis, Paul Lukowicz, Andrea Passarella, Alex Sandy Pentland, John Shawe-Taylor, Alessandro Vespignani

Artificial Intelligence Volume 339, February 2025, 104244

Read the full article at: www.sciencedirect.com