Biology subjects

Panda, N.

Publications and source records attributed to Panda, N..

5 recordsLinked to original sources

bioRxiv2026

The Geometry of Cognitive Difficulty: A Dynamical Manifold Theory in Excitable Neural Networks

Quantifying task difficulty remains an open theoretical problem in neuroscience and artificial intelligence. While difficulty is often treated as a scalar property of stimuli or optimization landscapes, neural computation unfolds as a transient reconfiguration of high-dimensional dynamical systems. Here we propose a dynamical manifold theory of difficulty based on heterogeneous, modular FitzHugh-Nagumo networks subjected to structured task demand. Task difficulty is modeled as a conflict-driven control parameter that perturbs competing neural submodules. We define four dynamical metrics: (i) transition action (energetic cost), (ii) peak dispersion entropy, (iii) coherence recovery deficit, and (iv) mean-field trajectory curvature. Across systematic sweeps of task demand, we demonstrate that difficulty does not collapse to a single axis but instead emerges as a multidimensional manifold. Energetic cost and dispersion entropy form a dominant axis, while geometric curvature and integration recovery exhibit partial independence and nontrivial correlations. These results suggest that cognitive difficulty corresponds to structured reorganization in neural state space rather than mere increases in activation amplitude. The proposed framework provides a biophysically interpretable foundation for linking neural dynamics, cognitive effort, and difficulty estimation in artificial systems.

biophysics↗

bioRxiv2025

Fingerprint-Based Explainable Machine Learning for Predicting Blood--Brain Barrier Permeability

Predicting blood-brain barrier (BBB) permeability is essential for early central nervous system (CNS) drug discovery, yet reliable computational screening remains challenging. This study presents a gradient-boosted ensemble framework trained on precomputed molecular fingerprints to classify compounds as BBB-permeable (BBB+) or non-permeable (BBB-). The 2048-bit fingerprints encode substructural information relevant to passive diffusion without requiring explicit physicochemical descriptors. The model, trained on experimentally annotated BBB datasets using Extreme Gradient Boosting (XGBoost) with Synthetic Minority Oversampling (SMOTE) to address class imbalance, achieved strong predictive performance (cross-validated ROC- AUC = 0.897 {+/-} 0.019; validation ROC-AUC = 0.932). External testing on literature-reported CNS-active compounds (Caffeine, Diazepam, Dopamine, and Levodopa) confirmed biological consistency: highly lipophilic drugs were predicted as BBB+, while polar molecules dependent on carrier-mediated transport were predicted as BBB-. The fingerprint-based model thus captures underlying permeability mechanisms through data-driven substructure learning. This approach eliminates the need for handcrafted descriptors while preserving interpretability through feature-importance analysis, establishing a reproducible, efficient, and explainable baseline for virtual BBB permeability screening in CNS drug development.

neuroscience↗

bioRxiv2025

Hybrid Epidemic--Neuronal Dynamics: A SEIR--FitzHugh--Nagumo Model for Information Flow in Complex Neural Networks

Information transfer in neural systems is often modeled through diffusive or synaptic mechanisms that fail to capture the contagion-like propagation of activation across large-scale networks. In this study, we introduce a hybrid SEIR-FitzHugh- Nagumo (FHN) model that integrates epidemiological dynamics with neuronal excitability to describe the flow of information through complex brain-like networks. Each node follows FHN excitability with slow recovery, while inter-node coupling obeys a modified SEIR process that regulates transmission probability based on exposure and recovery. This hybridization allows for the coexistence of oscillatory neural states and infection-like spreading modes, representing fast spiking communication constrained by population-level fatigue. We simulate the hybrid model across ring, Erd[o]s-Renyi, and Barabasi-Albert topologies and benchmark it against conventional diffusive FHN and FHN with synaptic depression (STD). Information-theoretic analysis using Mutual Information (MI) and Transfer Entropy (TE) shows that the hybrid system sustains higher directional information flow (TE-AUROC{approx} 0.52-0.54) and reduced latency across topologies. These results suggest that infection-inspired coupling enhances causal coherence and efficiency of information propagation in neural networks. The findings open a path toward multiscale hybrid models unifying epidemic, neuronal, and information-theoretic frameworks for understanding complex brain dynamics.

biophysics↗

bioRxiv2025

Quantum Tunneling-Gated Vesicle Fusion: Proton-Coupled Electron Transfer and Mechanical Barrier Softening Shape Neurotransmitter Release Latency

Neurotransmitter release in neurons requires synaptic vesicles to fuse rapidly with the presynaptic membrane after calcium entry, yet single-vesicle recordings show highly heterogeneous latency distributions with both fast events and long heavy tails. Most existing models fit these data empirically without mechanistic grounding. We introduce a quantum- mechanochemical model in which an initial proton-coupled electron transfer (PCET) step, governed by quantum tunneling, primes the vesicle for fusion, while a subsequent time-dependent mechanical barrier softening drives the final membrane merger. The scheme consists of three states: a closed SNARE complex that activates through PCET [Formula], a primed intermediate (P) that undergoes mechanical gating with an aging rate and forward rate k2(t), and a reversible slip-back process (k-1(t)) that sustains long-latency events. The PCET step is described by a Marcus-type tunneling expression with isotope-dependent mass terms, enabling direct prediction of the kinetic isotope effect (KIE) between protiated and deuterated conditions. Structural heterogeneity is included via a distribution of donor-acceptor distances. By calibrating the mechanochemical attempt frequency{gamma} to reproduce the typical early fusion probability (P [0- 5 ms]H{approx} 0.20), the model generates latency probability density functions (PDFs), cumulative distributions (CDFs), and hazard rates consistent with experimental observations. Parameter sweeps show how tunneling decay ({beta}tun) controls the KIE magnitude, while mechanical aging and back reaction redistribute early versus late events. This quantum-mechanical and force-activated framework provides a physically interpretable, testable alternative to purely empirical fits for single-vesicle fusion latency in neurons.

biophysics↗

bioRxiv2025

Controlling Neural Synchrony Through Variance-Driven Coupling in Complex Network Topologies

Adaptive control of synchrony in neuronal networks is central to understanding both normal brain function and pathological states such as epilepsy and tremor. We study a modified FitzHugh-Nagumo (FHN) network in which the local excitability is extended by a fifth-order nonlinearity and the global coupling strength adapts homeostatically to the spatial variance of neural activity. Using a combination of numerical bifurcation analysis and direct time-domain simulation, we map regimes of quiescence, stable fixed point, and self-sustained oscillation in the two-parameter space of input current and nonlinearity. The analytically predicted Hopf boundary agrees closely with the simulated transition to oscillations. Extending to networks with ring, Watts-Strogatz, and Barabasi-Albert topologies, we show that variance-driven adaptation strongly increases coupling and synchrony in rings, is partly suppressed in small-world graphs, and is almost ineffective in scale-free networks where hubs dominate connectivity. A simple feedback controller regulating the Kuramoto order parameter enables targeted desynchronization or resynchronization. Finally, stochastic forcing produces a non-monotonic impact on coherence, suggesting noise-induced resonance effects. This simulation-based framework links single-neuron excitability, network topology, and adaptive coupling control, and may inform strategies for brain-computer interfaces and neuromodulation therapies.

biophysics↗