Paper List
-
Pharmacophore-based design by learning on voxel grids
This paper addresses the computational bottleneck and limited novelty in conventional pharmacophore-based virtual screening by introducing a voxel cap...
-
CONFIDE: Hallucination Assessment for Reliable Biomolecular Structure Prediction and Design
This paper addresses the critical limitation of current protein structure prediction models (like AlphaFold3) where high-confidence scores (pLDDT) can...
-
On the Approximation of Phylogenetic Distance Functions by Artificial Neural Networks
This paper addresses the core challenge of developing computationally efficient and scalable neural network architectures that can learn accurate phyl...
-
EcoCast: A Spatio-Temporal Model for Continual Biodiversity and Climate Risk Forecasting
This paper addresses the critical bottleneck in conservation: the lack of timely, high-resolution, near-term forecasts of species distribution shifts ...
-
Training Dynamics of Learning 3D-Rotational Equivariance
This work addresses the core dilemma of whether to use computationally expensive equivariant architectures or faster symmetry-agnostic models with dat...
-
Fast and Accurate Node-Age Estimation Under Fossil Calibration Uncertainty Using the Adjusted Pairwise Likelihood
This paper addresses the dual challenge of computational inefficiency and sensitivity to fossil calibration errors in Bayesian divergence time estimat...
-
Few-shot Protein Fitness Prediction via In-context Learning and Test-time Training
This paper addresses the core challenge of accurately predicting protein fitness with only a handful of experimental observations, where data collecti...
-
scCluBench: Comprehensive Benchmarking of Clustering Algorithms for Single-Cell RNA Sequencing
This paper addresses the critical gap of fragmented and non-standardized benchmarking in single-cell RNA-seq clustering, which hinders objective compa...
Emergent Spatiotemporal Dynamics in Large-Scale Brain Networks with Next Generation Neural Mass Models
Universitat de les Illes Balears, Spain | Universitat Politècnica de Catalunya, Barcelona, Spain | Institut de Matemàtiques de la UPC - Barcelona Tech (IMTech), Barcelona, Spain | Centre de Recerca Matemàtica, Barcelona, Spain
The 30-Second View
IN SHORT: This work addresses the core challenge of understanding how complex, brain-wide spatiotemporal patterns emerge from the interaction of biophysically detailed local dynamics and empirical anatomical connectivity.
Innovation (TL;DR)
- Methodology Introduces a next-generation neural mass model (NG-NMM) into a large-scale brain network framework, providing a more biophysically grounded and analytically tractable description of population-level gamma oscillations via the PING mechanism.
- Methodology Applies the Master Stability Function (MSF) formalism and Floquet theory to systematically analyze transverse instabilities of homogeneous states (both fixed points and limit cycles) in a high-dimensional (90-node) network, linking instability modes to emergent spatiotemporal patterns.
- Biology Demonstrates that the network coupling in NG-NMMs enables cross-frequency coupling (CFC), specifically generating gamma oscillations whose amplitude is modulated by slower rhythms—a phenomenon not possible in isolated nodes and highly relevant for cognitive functions like memory.
Key conclusions
- NG-NMMs exhibit a broader dynamical repertoire than classical models, including regions of bistability, period-doubling cascades, and deterministic chaos within the homogeneous manifold (e.g., positive Lyapunov exponents for I_ext^E ~10-10.5 at ε=12).
- Anatomical connectivity is crucial for inducing cross-frequency coupling, allowing the emergence of gamma oscillations (27-170 Hz) with amplitude modulated by slower rhythms, a key feature of brain dynamics.
- The system's rich spatiotemporal patterns (traveling waves, high-dimensional chaos) arise from transverse instabilities of homogeneous solutions, analytically predicted by the MSF and confirmed via Lyapunov exponent and frequency spectrum analysis.
Abstract: Understanding the dynamics of large-scale brain models remains a central challenge due to the inherent complexity of these systems. In this work, we explore the emergence of complex spatiotemporal patterns in a large scale-brain model composed of 90 interconnected brain regions coupled through empirically derived anatomical connectivity. An important aspect of our formulation is that the local dynamics of each brain region are described by a next-generation neural mass model, which explicitly captures the macroscopic gamma activity of coupled excitatory and inhibitory neural populations (PING mechanism). We first identify the system’s homogeneous states—both resting and oscillatory—and analyze their stability under uniform perturbations. Then, we determine the stability against non-uniform perturbations by obtaining dispersion relations for the perturbation growth rate. This analysis enables us to link unstable directions of the homogeneous solutions to the emergence of rich spatiotemporal patterns, that we characterize by means of Lyapunov exponents and frequency spectrum analysis. Our results show that, compared to previous studies with classical neural mass models, next-generation neural mass models provide a broader dynamical repertoire, both within homogeneous states and in the heterogeneous regime. Additionally, we identify a key role for anatomical connectivity in cross-frequency coupling, allowing for the emergence of gamma oscillations with amplitude modulated by slower rhythms. These findings suggest that such models are not only more biophysically grounded but also particularly well-suited to capture the full complexity of large-scale brain dynamics. Overall, our study advances the analytical understanding of emerging spatiotemporal patterns in whole-brain models.