Michael W. Reimann

The result is a computationally cheap algorithm for connectivity at micro- meso- and macro-scale and a new view on the structure of connectivity. First diving into biological detail we emerged with a simplified model. Validated against the amazing MICrONS data.

We explain how morphologies work in terms of their constraints on connectivity, why these constraints are stronger than anticipated, and why they provide structure that other models don’t. Understanding this allowed us to capture all of this even without using morphologies.

How are we to understand the structure of synaptic connectivity at cellular resolution? Some say that predictions from neuronal morphologies can work well. Others point out that this provides limited insights into underlying mechanisms and is computationally too expensive.

Neuron morphological physicality and variability define the non-random structure of connectivity Connectivity in neuronal networks is characterized by high complexity that is required for the correct function of the circuitry. Our attempts to capture it has thus far been limited to the addition o...

My talk at CNS*2025 seemed to be well-received, so we wrote it up as a paper. doi.org/10.1101/2025.... Might be worth your time if you are interested in the wiring structure of neuronal circuitry - it's only 12 pages without methods.

Anyone at CNS*2025 in Florence? I hope to see you on Wednesday in our workshop: Bridging Complexity and Abstraction: Large-Scale Mechanistic Models of Brain Circuits from Biophysically Detailed to Simplified Representations.

Potentially even for inter-regional connectivity. Hope it will help with our understanding of neuronal connectivity. Readily usable for example in models and simulations of neuronal networks.

Much has been written about the structure and heterogeneity of local neuronal circuitry that is not captured by controls. We built a surprisingly simple graph based model and show that it recreates the complexity characterized in biological neuronal networks at different scales.

An algorithm to model the non-random connectome of cortical circuitry Neuronal connectivity has been characterized at various scales and with respect to various structural aspects. In models of connectivity, it has so far remained difficult to match all of them at once,...

Just released a preprint for everyone interested in connectomics. Specifically, local connectivity at cellular resolution, but also long-range connectivity. www.biorxiv.org/content/10.1...

I have embraced the complexity of the brain and data for a long time, and believe that approach is required. But that does not have to be overly centralized. Can be decentralized and incremental. Just focus on improving existing resources instead of reinventing the wheel.

It's finally out!

Visual experience orthogonalizes visual cortical responses

Training in a visual task changes V1 tuning curves in odd ways. This effect is explained by a simple convex transformation. It orthogonalizes the population, making it easier to decode.

10.1016/j.celrep.2025.115235

Does the culture you grow up in shape the way you see the world? In a new Psych Review paper, @chazfirestone.bsky.social & I tackle this centuries-old question using the Müller-Lyer illusion as a case study. Come think through one of history's mysteries with us🧵(1/13):

Leslie Ungerleider (1946 - 2020) was an extraordinary, pioneering neuroscientist. The latest issue of @jocn.bsky.social honors her legacy with articles authored by former colleagues & trainees that highlight critical aspects of her work and its influence https://buff.ly/4156UKw #cogsci #neuroscience

Maybe I read too much into the two sentences, but it seems to imply something like: You plug Bernoulli's principle into a computer, then optimize for agent mobility and end up with a weird looking quadcopter. Conclusion: Clearly birds are flying using four rotors.

petition to change the word describing ChatGPT's mistakes from 'hallucinations' to 'confabulations'

A hallucination is a false subjective sensory experience. ChatGPT doesn't have experiences!

It's just making up plausible-sounding bs, covering knowledge gaps. That's confabulation

Spike sorting biases and information loss in a detailed cortical model Sorting electrical signals (spikes) from extracellular recordings of large groups of connected neurons is essential to understanding brain function. Despite transformative advances in dense extracellu...

Lots of neuroscience discoveries are based on spike sorting data, but we know it captures only parts of what's going on. Through modeling we predict which parts are captured and which are missed. 🧪

We hope the data set helps improve spike sorting even further!

www.biorxiv.org/content/10.1...

Frontiers | Data-Driven Modeling of Cholinergic Modulation of Neural Microcircuits: Bridging Neurons, Synapses and Network Activity Neuromodulators, such as acetylcholine (ACh), control information processing in neural microcircuits by regulating neuronal and synaptic physiology. Computat...

www.biorxiv.org/content/10.1...

www.frontiersin.org/journals/neu...

Michael W. Reimann ‪Group Leader @ Blue Brain, EPFL, Switzerland‬ - ‪‪Cited by 3,298‬‬ - ‪cortical circuits‬ - ‪modeling‬ - ‪connectomics‬ - ‪neural codes‬

Hello Monica @pkpd-babe.bsky.social . Could you please add me to the science feed? I am a neuroscientist working on connectomics with simulation-based methods. Scholar profile: scholar.google.com/citations?us...

🤔 Yeah that makes a lot of sense to me.
I like this way of thinking in the context of constraints, in this case: the imperative to conserve energy. Needed to understand the brain, in my opinion.

I would ask to what degree the current parcellation into brain regions works for describing the rules of the mapped connections. Can we improve it further?

But properly manipulating connectivity in detailed models is harder than it sounds. Changing one aspect while keeping others fixed is tricky due to complex dependencies. Our tool helps with this. We show examples and results of how this plays out in practice.

Ultimately, this can only be tested through manipulations that are currently only possible in simulation. Due to the wealth of information EM gives us about connections (dendritic locations, sizes of synapses, types of participating neurons) we would do this in very detailed models.

The motivation for the work came from ongoing and published electron microscopic connectome reconstructions that were said to solve the connectomics of brain microcircuitry. But even knowing all connections, one can describe any number of rules and trends - which ones are functionally meaningful?

A connectome manipulation framework for the systematic and reproducible study of structure–function relationships through simulations Abstract. Synaptic connectivity at the neuronal level is characterized by highly non-random features. Hypotheses about their role can be developed by correlating structural metrics to functional featu...

I am happy to announce that our latest work has been published in The MIT Press' Network Neuroscience (doi.org/10.1162/netn...). In this methods paper, we present a novel Python framework for rapid connectome manipulations of detailed, biologically realistic network models in SONATA format.

How the Web of Science takes a step back <p>The Web of Science, a major commercial indexing service of scientific journals operated by Clarivate, recently decided to remove eLife from its Science Citation Index Expanded (SCIE). eLife will on...

Great blog by Bodo Stern (@hhmi.bsky.social) about the recent Clarivate decision on @elife.bsky.social’s impact factor. www.coalition-s.org/blog/how-the...

Our SSCx model (linktr.ee/BlueBrainPjt) provided a valuable null model to compare findings against: It has the excitatory “fan-out tracks”, but not the specific inhibition. Clearly: Inhibition is computationally powerful and should be studied more! (3/3)

A schematic that describes the main excitatory + inhibitory motif encountered in the data.

We found that the higher-order structure of excitation contains specific motifs that form a backbone of connectivity in a fan-out structure. Inhibitory connectivity is structured by this, providing specific competition between the motifs. Disinhibition in turn targets these neurons (2/3)

Specific inhibition and disinhibition in the higher-order structure of a cortical connectome Abstract. Neurons are thought to act as parts of assemblies with strong internal excitatory connectivity. Conversely, inhibition is often reduced to blanke

We are happy to announce that our latest connectomics work has been published by Cerebral Cortex. We have analyzed one of our favorite open datasets: MICrONS cortical mm^3. In this EM dataset we found new rules for neuronal connectivity & describe them (doi.org/10.1093/cerc...) (1/3)

My favorite sentence from the paper: "Moreover, ANNs usually disregard the fact that biological networks are spatially
embedded networks running under constraints, a feature we discussed earlier as a potential key to understanding
neural network organization."
Could not agree more!