Stories
16 stories, grouped by topic · 한국어 글 (Korean posts)
Cells & wiring
What Fraction of Neurons Are Inhibitory in the Human Cortex? About 31% in Our Count (vs 12% in Mouse)
In about one cubic millimetre of human cortex (H01), 4,688 of 15,219 classified neurons are inhibitory interneurons: 31%. In a mouse cortex map (MICrONS) we count 12%. How we counted, how it changes by layer, and what the number does not tell you.
What Is in Layer 1 of the Human Cortex? Few Neurons, Many Astrocytes
In a cubic millimetre of human cortex (H01), only 18% of layer-1 cells are neurons (827 of 4,595), against 70% in layer 2. The most common cell is the astrocyte: 37%. What else is there, why the layer is mostly wiring, and the caveats.
Which Brain Cells Spread Out Evenly? Microglia and OPCs in Every Cortical Layer
In a cubic millimetre of human cortex (H01), microglia and OPCs sit at about 6,000 per mm³ in every layer, from 5,854 to 6,519 (a 1.1-fold range). Neurons vary 6.4-fold across the same layers. The numbers, one likely reason, and the caveats.
What Is the Most Common Cell in the Human Brain? In This Sample, Not the Neuron
In about one cubic millimetre of human cortex mapped with an electron microscope (H01), oligodendrocytes outnumber neurons 20,139 to 16,087. Half of them sit in the white matter. In the grey matter alone, neurons still win. Here is the count, layer by layer.
Why Is White Matter White? 31.5% of Its Mapped Volume Is Myelinated Axons
White matter looks white because it is packed with myelin, a fatty wrapping around nerve fibres. In a real human cortex sample (H01), myelinated axons fill 31.5% of the mapped white-matter volume but only 1.0% of layer 1. Raw electron-microscope images and the numbers.
How Close Are Neurons to Blood Vessels? We Measured 10,441 of Them in a Real Human Brain Sample
In about one cubic millimetre of human cortex mapped with an electron microscope (H01), 98% of neurons sit within 50 µm of a blood vessel (median 23 µm). How we measured it, why neurons stay so close, and what the number does not tell you.
Human vs Mouse vs Fly: How We Measured Five Rounds of Brain-Cell Comparisons (and What We Got Wrong First)
We compared human, mouse and fruit-fly neurons in five rounds using three public connectomes. This post lists every number, how it was measured, the seven mistakes we caught before publishing, and why the comparison is not a fair fight.
We Followed One Neuron Through a Map of the Human Brain — Methods, Numbers, and What We Got Wrong
We opened the public H01 human-cortex connectome and counted every input to a single layer-5 pyramidal neuron: 6,131 synapses from 4,761 axon segments. Here are the methods, the figures, and the caveats our video could only hint at.
Drugs & receptors
Is Dopamine the Pleasure Chemical? Not Quite. 390 Fly Neurons, One Receptor, and What the Experiments Say
We counted the predicted dopamine neurons in a complete fruit-fly connectome, looked at dopamine bound to the human D1 receptor, and checked the 'pleasure chemical' label against the monkey and rat experiments behind it. Includes the four things we fixed before publishing.
Why Do Some Allergy Pills Make You Sleepy? Eight Histamine Neurons, One Receptor Pocket, and a Door
We counted the histamine neurons in a complete fruit-fly connectome, overlaid two real antihistamine structures on the human H1 receptor, and followed the evidence for why only some allergy pills make you drowsy. Includes the two counting mistakes we caught before publishing.
Same Lock, Different Effects: Why LSD Is Psychedelic and Lisuride Isn't
LSD and lisuride occupy the same pocket of the human serotonin 2A receptor, 0.12 Å apart. Only LSD is psychedelic. We compared five published crystal structures residue by residue, checked the LSD pose against an independent structure from another lab, and went through what the literature says explains the difference.
Why Is Nicotine So Hard to Quit? What the Real Receptor Structures Show
We overlaid two published cryo-EM structures of the human α4β2 nicotinic receptor, one holding nicotine and one holding acetylcholine. In all three pockets the two molecules sit about 1 Å apart. Here is what that shows, and what it doesn't.
Myths checked
Does Parkinson's Start Only After 80% of Neurons Die? We Traced the Number Back to Its Sources
A widely repeated line says 80% of dopamine neurons are already gone when Parkinson's symptoms appear. We read the studies behind it. For the cells themselves, the best estimate is closer to 30%; the 80% is about something else.
Methods & corrections
Which H01 Is It? Two Public Synapse Releases, Only One Matches the Paper — and How We Found Out the Hard Way
The public H01 human-cortex bucket holds two synapse layers. 20210601 is raw detection output; 20210729 matches the paper's per-cell Table S14 exactly. We used the wrong one in two posts. Here is how to tell them apart.
Does Chronic Stress Really Lower Hippocampal BDNF? What an LLM Read of 9,585 Studies Actually Says
A plain-language companion to our preprint: an LLM extracted structured data from 9,585 stress–brain studies and found hippocampal BDNF reported down and up about equally, neuroinflammation overtaking the HPA axis since ~2018, and a structural-vs-functional split behind the amygdala contradiction.
Is DRD2 Really a Depression Drug Target? How Colocalization Demoted the Top Signal and Put SLC12A5 (KCC2) Forward
A plain-language companion to our preprint: the strongest TWAS signal for major depressive disorder (DRD2) failed colocalization, so we demoted dopaminergic drugs to an exploratory tier and re-anchored on SLC12A5 (KCC2) — the one gene that passed SMR, HEIDI, and coloc across 635 GWAS loci.
Brain health (archive)
Earlier evidence summaries on sleep, stress, memory, supplements and dementia. They are kept as they were and are not part of the connectome series.
Brain Supplements: Evidence-Based Guide to What Actually Works (2026)
Omega-3, magnesium L-threonate, lion's mane, ginkgo, NAD+ precursors — a neuroscience-backed evaluation of brain supplements sorted into four evidence tiers (Strong/Some/Weak/Pseudo). Includes dosing, drug interactions, age-specific recommendations, and which products are pure marketing vs which actually have RCT support.
Dementia vs Normal Aging: MMSE/MoCA Cognitive Tests + Brain Imaging Guide 2026
How to tell normal age-related forgetfulness from early dementia — neurological differences, MMSE and MoCA scoring with cutoffs, MRI/CT/PET imaging guide, 2026 breakthrough blood biomarkers (p-tau217), and a practical checklist for when to see a neurologist.
Sleep Deprivation Effects on the Brain: A 2026 Neuroscience Guide
Sleep loss isn't just tiredness. It slows glymphatic clearance, weakens the hippocampus's ability to encode new memories, makes the amygdala overreact, and raises brain amyloid-β. From one all-nighter to chronic sleep debt, this guide covers what verified studies actually show.
How Chronic Stress Shrinks Your Hippocampus: What Neuroscience Actually Shows
Chronic stress physically shrinks the hippocampus and hyperactivates the amygdala. Here's the neuroscience behind how stress rewires your brain — and the evidence-based strategies to reverse the damage.
Cortisol and Memory: Complete Neuroscience Guide (2026 Update)
How does cortisol level change memory? Acute stress can strengthen memory, while chronic stress impairs it. The neuroscience of the cortisol-memory relationship — the inverted-U curve, hippocampal changes, ageing — and practical ways to protect memory.
Does Meditation Rewire Your Brain? What fMRI and MRI Studies Actually Show
Functional (fMRI, EEG) and self-report effects of meditation are reasonably consistent. Structural claims are not: the largest randomized trial found no brain-structure change after 8 weeks of MBSR. Here is what each key study actually measured.
Neuroscience of Addiction Explained — How the Brain Gets Hijacked
Understand addiction through neuroscience. Learn how drugs and behaviors hijack the brain's reward system, why willpower isn't enough, and what modern treatments offer.