Biology subjects

Khona, M.

Publications and source records attributed to Khona, M..

6 recordsLinked to original sources

bioRxiv2024

A morphing map model for place field organization in large environments

Navigation and spatial memory, essential for animal survival, are founded on the hippocampal formations ability to host spatial neurons marking an animals position. Within controlled settings, the place cells of the hippocampus are activated in specific regions, the place fields. In this paper, we introduce a "morphing map algorithm" based on bat data from a 200-meter tunnel, highlighting a multifield, multiscale spatial representation of place fields. This algorithm posits that in vast environments, bats opt to select landmarks or anchors for spatial representation. The method introduces a hyperbolic geometry in spatial representation, aligning with bat data from vast areas. We apply Procrustes distance to evaluate similarities between neural maps of bats and model-generated maps, revealing shared structure across bat maps, which might be dictated by common anchor points. Further explorations into anchor point ablation provide insights into their crucial role in map stability. The results present an evolved neural navigation strategy in expansive habitats and guide future spatial representation research across species.

neuroscience↗

bioRxiv2024

Self-organized emergence of modularity, hierarchy, and mirror reversals from competitive synaptic growth in a developmental model of the visual pathway

A hallmark of the primate visual system is its architectural organization consisting of multiple distinct (modular) areas that connect hierarchically. These areas exhibit specific spatial organization on the cortical sheet, with primary visual cortex at the center and subsequent regions in the hierarchy encircling the earlier one, and detailed topological organization, with retinotopy in each area but striking mirror reversals across area boundaries. The developmental rules that drive the simultaneous formation of these architectural, spatial, and topographic aspects of organization are unknown. Here we demonstrate that a simple synaptic growth rule driven by spontaneous activity and heterosynaptic competition generates a detailed connectome of the visual pathway, with emergence of all three types of organization. We identify a theoretical principle -- local greedy wiring minimization via spontaneous drive (GWM-S) -- implemented by the mechanism, and use this insight to propose biologically distinct growth rules that predict similar endpoints but testably distinguishable developmental trajectories. The same rules predict how input geometry and cortical geometry together drive emergence of hierarchical, convolution-like, spatially and topographically organized sensory processing pathways for different modalities and species, providing a possible explanation for the observed pluripotency of cortical structure formation. We find that the few parameters governing structure emergence in the growth rule constitute simple knobs for rich control, that could (potentially genetically) encode a projection neuron-like connectivity patterns and interneuron-like ones. In all, the presented rules provide a parsimonious mechanistic model for the organization of sensory cortical hierarchies even without detailed genetic cues for features like map reversal, and provide numerous predictions for experiment during normal and perturbed development.

neuroscience↗

bioRxiv2022

No Free Lunch from Deep Learning in Neuroscience: A Case Study through Models of the Entorhinal-Hippocampal Circuit

Research in Neuroscience, as in many scientific disciplines, is undergoing a renaissance based on deep learning. Unique to Neuroscience, deep learning models can be used not only as a tool but interpreted as models of the brain. The central claims of recent deep learning-based models of brain circuits are that they make novel predictions about neural phenomena or shed light on the fundamental functions being optimized. We show, through the case-study of grid cells in the entorhinal-hippocampal circuit, that one may get neither. We begin by reviewing the principles of grid cell mechanism and function obtained from first-principles modeling efforts, then rigorously examine the claims of deep learning models of grid cells. Using large-scale architectural and hyperparameter sweeps and theory-driven experimentation, we demonstrate that the results of such models may be more strongly driven by particular, non-fundamental, and post-hoc implementation choices than fundamental truths about neural circuits or the loss function(s) they might optimize. We discuss why these models cannot be expected to produce accurate models of the brain without the addition of substantial amounts of inductive bias, an informal No Free Lunch result for Neuroscience. Based on first principles work, we provide hypotheses for what additional loss functions will produce grid cells more robustly. In conclusion, circumspection and transparency, together with biological knowledge, are warranted in building and interpreting deep learning models in Neuroscience.

neuroscience↗

bioRxiv2022

A taste for numbers: Caenorhabditis elegans foraging follows a low-dimensional rule of thumb

Rules of thumb are behavioral algorithms that approximate optimal behavior while lowering cognitive and sensory costs. One way to reduce these costs is by reducing dimensionality: While the theoretically optimal behavior may depend on many environmental variables, a rule of thumb may use a low-dimensional combination of variables that performs reasonably well. Experimental proof of a dimensionality reduction requires an exhaustive mapping of all relevant combinations of several environmental parameters, which we performed for Caenorhabditis elegans foraging by covering all combinations of food density (across 4 orders of magnitude) and food type (across 12 bacterial strains). We found a one-dimensional rule: Worms respond to food density measured as number of bacteria per unit surface, disregarding other factors such as biomass content or bacterial strain. We also measured fitness experimentally, determining that the rule is near-optimal and therefore constitutes a rule of thumb that leverages the most informative environmental variable.

animal behavior and cognition↗

bioRxiv2021

Spontaneous emergence of topologically robust grid cell modules: A multiscale instability theory

Modular structure and function are ubiquitous in biology, from the organization of animal bodies and brains to the scale of ecosystems. However, the mechanisms of modularity emergence remain unclear. Here we introduce the principle of peak selection, a process by which purely local interactions and smooth gradients can result in global modular organization. It can lead to the self-organization of discontinuous module boundaries from a smooth global gradient, unifying the positional hypothesis and the Turing pattern formation hypothesis for morphogenesis. Applied to the brain's grid cell networks, peak selection results in the spontaneous emergence of functionally distinct modules with discretely spaced spatial periods. Applied to ecological systems, a generalization of the process results in discrete systems-level niches. The dynamics exhibits emergent self-scaling to variations in system size and "topological robustness" that renders module emergence and module properties insensitive to most parameters. Further, peak selection confers robustness within modules. It ameliorates the fine-tuning requirement of continuous attractor dynamics even in single grid cell modules. It makes a detail-independent prediction that grid module period ratios should approximate adjacent integer ratios, furnishing the most accurate match to data to date. Additional testable predictions promise to bridge physiology, connectomics, and transcriptomics. In sum, our results indicate that local interactions combined with low-information global gradients can drive robust global module emergence.

neuroscience↗

bioRxiv2020

Reverse-engineering Recurrent Neural Network solutions to a hierarchical inference task for mice

We study how recurrent neural networks (RNNs) solve a hierarchical inference task involving two latent variables and disparate timescales separated by 1-2 orders of magnitude. The task is of interest to the International Brain Laboratory, a global collaboration of experimental and theoretical neuroscientists studying how the mammalian brain generates behavior. We make four discoveries. First, RNNs learn behavior that is quantitatively similar to ideal Bayesian baselines. Second, RNNs perform inference by learning a two-dimensional subspace defining beliefs about the latent variables. Third, the geometry of RNN dynamics reflects an induced coupling between the two separate inference processes necessary to solve the task. Fourth, we perform model compression through a novel form of knowledge distillation on hidden representations - Representations and Dynamics Distillation (RADD)- to reduce the RNN dynamics to a low-dimensional, highly interpretable model. This technique promises a useful tool for interpretability of high dimensional nonlinear dynamical systems. Altogether, this work yields predictions to guide exploration and analysis of mouse neural data and circuity.

neuroscience↗