Systems Neuroscience and Computational Modeling

Vocabulary practice flashcards covering core concepts, methodologies, microcircuits, graph theory, metabolic modeling, and machine learning architectures from Systems Neuroscience.

Systems Neuroscience

The branch of neuroscience that studies how populations of neurons and neural circuits produce perception, cognition, and behavior, bridging the gap between molecular/cellular neuroscience and cognitive neuroscience through emergent properties rather than a purely reductionist approach.

Integrative Systems Neuroscience

The union of systems biology and integrative neuroscience that incorporates interdisciplinary and multiscale analysis of nervous systems, utilizing computational models to evaluate how different neural components interact.

Neuroimaging Resolution Landscape

The systematic mapping of neuroscience methodologies based on spatial resolution (meters) and temporal resolution (seconds), ranging from high-temporal electrophysiology (penetrating microelectrodes, ECoG, EEG, MEG) to high-spatial structural and functional imaging (MRI, fMRI, PET).

Calcium Imaging

An invasive imaging technique that tracks neuronal activity by measuring changes in intracellular calcium ions using fluorescent indicators, offering cellular-level spatial resolution.

Haemodynamic Response

A local change in blood flow and oxygenation associated with neural activity, measured via functional imaging techniques (such as fMRI) due to differences in light and magnetic absorption between oxygenated and deoxygenated blood.

Guanidinoacetate (GAA) Transferase Deficiency

An enzymatic deficiency that prevents proper creatine synthesis and transport in the brain, depleting cellular energy reserves needed for ATP storage and resulting in neuronal misfiring and epilepsy.

Epigenetic Methylation

A chemical modification that frequently occurs at cytosine-guanine ($\text{C-G}$) sites, reducing the likelihood of RNA polymerase binding and consequently lowering gene expression.

Histone Acetylation

The addition of an acetyl group to a positively charged histone protein, rendering it negatively charged and releasing it from DNA to facilitate transcription.

Genome-Wide Association Study (GWAS)

An observational approach that scans evenly spaced genetic proxies across the genome to detect single-nucleotide variants statistically associated with a phenotype or disease trait above a genome-wide significance threshold.

Steady-State Assumption in Metabolic Modeling

The core condition in constraint-based metabolic modeling assuming metabolite concentrations do not change over time, mathematically expressed by the stoichiometric matrix ($S$) and flux vector ($V$) as $S \times V = 0$.

Context-Specific Metabolic Model

A refined metabolic network model tailored to a specific cell type, tissue, or condition by using omics data (e.g., transcriptomics) to constrain unexpressed reactions to zero flux and removing inactive pathways.

Bottom-Up Modeling

A modeling paradigm that starts with individual biological components and measurements (neuronal, synaptic) and combines them to predict macroscale system effects and outputs.

Top-Down Modeling

A modeling approach addressing inverse problems by constraining the desired output or behavioral outcome and working backward to infer potential underlying mechanisms.

Neuron Anatomy

The structural composition of a nerve cell, including dendrites, cell body (soma), nucleus, axon hillock, myelin sheath, nodes of Ranvier, Schwann cells, and axon terminals.

Excitatory Postsynaptic Potential (EPSP)

A local depolarization of the postsynaptic membrane that increases the membrane potential, making the neuron more likely to generate an action potential.

Inhibitory Postsynaptic Potential (IPSP)

A local hyperpolarization or hypopolarization of the postsynaptic membrane, predominantly mediated by neurotransmitters such as GABA and glycine, making the neuron less likely to fire an action potential.

Microcircuit Motifs

Recurring organizational wiring patterns among groups of neurons, including feedforward excitation, feedforward inhibition, convergence/divergence, lateral inhibition, feedback/recurrent inhibition, and recurrent excitation.

Lateral Inhibition

A microcircuit mechanism in which an excited neuron suppresses the activity of its neighboring parallel neurons, serving to sharpen receptive boundaries and enhance sensory contrast.

Recurrent Excitation

A circuit motif in which an activated neuron provides positive excitatory feedback to itself or interconnected neighboring cells, supporting signal amplification and memory consolidation.

Canonical Microcircuit

A stereotypical, multi-layered cortical circuit architecture (layers L1 through L6) where granular layer 4 receives sensory thalamic input and distributes signals to supragranular (L2/3) and infragranular (L5/6) layers.

Hippocampal Trisynaptic Circuit

A microcircuit sequence critical for learning and memory wherein cortical information enters via the perforant path into the dentate gyrus, projects via mossy fibers to CA3, and continues via Schaffer collaterals to CA1.

Stretch Reflex (Patellar Reflex)

A monosynaptic and polysynaptic spinal reflex arc where tapping the patellar tendon stretches muscle spindles, activating sensory afferent axons that directly excite agonist alpha motor neurons while activating inhibitory interneurons to suppress antagonist motor neurons.

Excitation/Inhibition (E/I) Balance

The steady-state ratio between excitatory and inhibitory inputs within a neural circuit; loss of this balance can cause runaway excitation resulting in seizures, epilepsy, or cognitive and social behavioral deficits.

Optogenetics

A biological tool combining genetics and optics to express light-sensitive proteins (such as step-function opsins, SSFO) in specific cell types, enabling millisecond-scale causal control of neuronal firing using light.

Connectomics

The subfield of neuroscience dedicated to mapping and analyzing the complete network of structural, functional, and effective connections in the brain.

Structural Connectivity

The anatomical wiring of the nervous system representing physical biological pathways, such as axonal projections and white-matter fiber tracts mapped via tract tracing or diffusion MRI.

Functional Connectivity

The statistical correlation or dependency of physiological activity between distinct, spatially separated brain areas across time.

Effective Connectivity

The directed, causal influence that one neuronal group or brain region exerts over another, establishing the directionality of signal flow.

Connectome Connectivity Matrix

A 2D matrix representation of macroscale brain wiring where rows and columns indicate cortical source and target areas, displaying symmetric entries for undirected graphs and asymmetric entries for directed connections.

Modularity

A graph-theoretic metric quantifying the degree to which a network is organized into segregated, densely interconnected communities or functional modules.

Centrality

A topological network measurement determining the relative structural or functional importance of a specific node (e.g., degree centrality, identifying network hubs).

Clustering Coefficient

The probability that two neighboring nodes connected to a common third node are also connected directly to one another, reflecting the degree of local specialization in a network.

Shortest Path Length

The minimum number of edges or intermediate steps needed to navigate between two nodes in a network, quantifying the efficiency of global information integration.

Small-World Network

A network topology characterized simultaneously by high local clustering and short average path length, balancing segregated local processing with rapid global integration.

Supervised Learning

A machine learning paradigm where an algorithm is provided with inputs paired with ground-truth target labels to learn a predictive mapping function.

Unsupervised Learning

A machine learning paradigm where an algorithm identifies inherent hidden patterns, clusters, or structures within unlabeled data without external target feedback.

Reinforcement Learning

A machine learning framework where an agent interacts dynamically with an environment, learning an optimal behavioral policy through numerical reward and punishment signals.

Overfitting and Underfitting

Model fitting failures in classification and regression: overfitting occurs when a model memorizes training data and noise at the expense of generalization, while underfitting occurs when a model lacks sufficient complexity to capture underlying data trends.

Deep Neural Network Architecture

A computational model composed of stacked node layers—including an input layer, multiple hidden processing layers, and an output layer—connected via adjustable weights and biases.

Activation Function

A mathematical operation applied to the weighted sum of a neural network unit's inputs that introduces non-linearity and mirrors biological neuronal firing response curves.

Rectified Linear Unit (ReLU)

A piecewise linear activation function defined as $f(x) = \max(0, x)$ that sets all negative inputs to zero and leaves positive inputs unchanged, widely used to prevent saturation in deep networks.

Recurrent Neural Network (RNN)

A neural network architecture featuring internal cyclical connections, enabling the network to maintain an internal state or memory of prior sequential inputs.

Neural Mass Model

A computational model that simulates the collective average firing rate and membrane potential dynamics of large neuronal populations rather than tracking individual single cells.

Neural Field Model

A continuous computational modeling framework that incorporates continuous two-dimensional or three-dimensional spatial coordinates to evaluate spatiotemporal activity across cortical sheets.

Representational Similarity Analysis (RSA)

A systems and computational neuroscience framework that compares geometric representational structures between biological neural activity patterns (e.g., fMRI or multi-unit arrays) and artificial neural network layer representations.