tooluniverse-neuroscience

# Neuroscience Research Skill

**KEY PRINCIPLES**: LOOK UP, DON'T GUESS — use PubMed/EuropePMC for neuroanatomy facts, WormBase for C. elegans connectome data, UniProt for neural protein properties. Verify claims with literature before answering. Use Python computation for quantitative neuroscience problems.

---

## LOOK UP, DON'T GUESS
When uncertain about any neuroscience fact — brain region function, neural circuit connectivity, ion channel properties, neurotransmitter receptor subtypes — SEARCH databases first. A PubMed-verified answer is always more reliable than reasoning from memory. This is especially critical for neuroanatomy, where structures have precise boundaries and connectivity patterns that are easy to confuse.

---

## 1. Computational Neuroscience Reasoning

### Rate-Based Models
- Firing rate of a neuron: r = f(I - theta), where I = total synaptic input, theta = threshold, f = transfer function (sigmoid, ReLU, or threshold-linear)
- Balanced excitation/inhibition: in cortical networks, excitatory and inhibitory inputs are large but nearly cancel, leaving a small net drive
- Population rate equations: tau * dr/dt = -r + f(W*r + I_ext), where W = connectivity matrix
- Steady-state analysis: set dr/dt = 0, solve r = f(W*r + I_ext) — use fixed-point iteration or Newton's method

### Integrate-and-Fire Neurons
- Membrane voltage dynamics: tau_m * dV/dt = -(V - V_rest) + R_m * I(t)
- When V reaches threshold V_th: emit spike, reset to V_reset, enter refractory period tau_ref
- Firing rate for constant input: r = 1 / (tau_ref + tau_m * ln((R_m*I - V_reset) / (R_m*I - V_th))) [valid when R_m*I > V_th]
- For sub-threshold input: neuron requires fluctuations (noise) to fire — noise-driven regime
- Key variants: LIF (leaky), EIF (exponential), AdEx (adaptive exponential), Izhikevich (2D with recovery variable)

### Synaptic Plasticity
- **STDP** (Spike-Timing-Dependent Plasticity):
  - Pre-before-post (positive dt): LTP (potentiation) — synapse strengthened
  - Post-before-pre (negative dt): LTD (depression) — synapse weakened
  - Window shape: typically exponential decay with tau_+ ~ 20ms (LTP) and tau_- ~ 20ms (LTD)
- **Hebbian learning**: "cells that fire together wire together" — correlation-based; unstable without normalization
- **BCM theory**: sliding threshold — low postsynaptic activity → LTD, high → LTP; threshold slides with average activity
- **Homeostatic plasticity**: synaptic scaling adjusts all synapses multiplicatively to maintain target firing rate

### Network Dynamics
- **Mean-field theory**: replace individual neurons with population-averaged firing rates; self-consistency equation r = f(J*r*sqrt(K) + I_ext) where K = number of connections
- **Balanced networks**: E/I balance emerges when sqrt(K)*J ~ O(1); firing rate ~ (mu - theta) / tau where mu = mean input, theta = threshold
- **Chaos transition**: in random networks, chaos onset at g_c = 1 (gain parameter); above g_c, autocorrelation decays, Lyapunov exponent > 0
- **Oscillations**: gamma (30-80 Hz) from E-I loops (PING model), theta (4-8 Hz) from slower inhibition or hippocampal circuits, alpha (8-12 Hz) from thalamo-cortical loops

### Quantitative Problem-Solving Strategy
1. Identify the model type (single neuron, network, plasticity rule)
2. Write down the governing equations with all parameters
3. **ALWAYS use Python** for multi-step calculations — do not attempt mental arithmetic
4. Check units: voltages in mV, currents in nA or pA, time constants in ms, rates in Hz
5. Sanity check: cortical firing rates are typically 1-20 Hz; tau_m ~ 10-20 ms; V_th ~ -50 mV

---

## 2. Neuroanatomy Reasoning

### CRITICAL: Look Up Neuroanatomy
Brain region functions, boundaries, and connectivity are precise anatomical facts. When asked about specific regions, nuclei, or tracts:
1. Search PubMed or EuropePMC with specific anatomical terms
2. For connectivity: search "[region A] projection [region B]" or "[region] afferents efferents"
3. For function: search "[region] lesion" or "[region] function review"

### Human Brain — Major Divisions
- **Cerebral cortex**: frontal (motor, executive), parietal (somatosensory, spatial), temporal (auditory, memory), occipital (visual)
- **Basal ganglia**: caudate + putamen (striatum) → GPi/SNr (output) → thalamus; direct pathway (facilitate movement) vs indirect pathway (suppress movement); dopamine from SNc modulates both
- **Cerebellum**: coordination, timing, motor learning; receives mossy fibers (pontine nuclei) and climbing fibers (inferior olive); Purkinje cells are sole output of cerebellar cortex
- **Brainstem**: midbrain (superior/inferior colliculi, substantia nigra, red nucleus), pons (pontine nuclei, respiratory centers), medulla (cardiovascular/respiratory centers, cranial nerve nuclei)
- **Thalamus**: relay station — every sensory modality (except olfaction) synapses here before cortex; also receives cortical feedback (corticothalamic loops)
- **Hippocampus**: declarative memory formation; trisynaptic circuit: EC → DG → CA3 → CA1 → EC; place cells, grid cells

### Model Organism Neuroanatomy
- **C. elegans**: 302 neurons, complete connectome mapped; use `WormBase_get_gene` for gene expression, neuron identity, connectivity data
- **Drosophila**: mushroom body (learning/memory), antennal lobe (olfaction), central complex (navigation); ~100,000 neurons; FlyWire connectome
- **Zebrafish**: transparent larvae for whole-brain imaging; Mauthner cells (escape response); use `Alliance_search_genes` for orthologs
- **Mouse**: Allen Brain Atlas for gene expression; use PubMed for circuit tracing studies (rabies virus, optogenetics)

### Reasoning Pattern for "Where in the Brain?" Questions
1. Identify the function asked about (motor, sensory, memory, emotion, language)
2. Map to candidate regions from general knowledge
3. VERIFY with PubMed search: "[function] brain region fMRI" or "[function] lesion study"
4. Check for lateralization (language → usually left hemisphere)
5.