Sam Mestern

Where in the brain does DANDI have data? Now you can find out. DANDI Atlas maps datasets onto the Allen CCF mouse brain atlas in interactive 3D. Explore 47 dandisets across 353 brain regions. atlas.dandiarchive.org Announcement: about.dandiarchive.org/blog/2026/02/2… youtube.com/watch?v=D8514C…

Today, I’m launching DAAF, the Data Analyst Augmentation Framework: an open-source, extensible workflow for Claude Code that allows skilled researchers to rapidly scale their expertise and accelerate data analysis by as much as 5-10x -- without sacrificing the transparency, rigor, or reproducibility demanded by our core scientific principles. You (yes, YOU!) can install and begin using it in as little as 10 minutes from a fresh computer with a high-usage Anthropic account (crucial accessibility caveat, unfortunately very expensive!) github.com/DAAF-Contribut… DAAF explicitly embraces the fact that LLM-based research assistants will never be perfect and can never be trusted as a matter of course. But by providing strict guardrails, enforcing best practices, and ensuring the highest levels of auditability possible, DAAF ensures that LLM research assistants can still be immensely valuable for critically-minded researchers capable of verifying and reviewing their work. In energetic and vocal opposition to deeply misguided attempts to replace human researchers, DAAF is intended to be a force-multiplying "exo-skeleton" for human researchers (i.e., firmly keeping humans-in-the-loop). The base framework comes ready out-of-the-box to analyze any or all of the 40+ foundational public education datasets available via the Urban Institute Education Data Portal (educationdata.urban.org/documentation/), and is readily extensible to new data domains and methodologies with a suite of built-in tools to ingest new data sources and craft new Skill files at will! With DAAF, you can go from a research question to a shockingly nuanced research report with sections for key findings, data/methodology, and limitations, as well as bespoke data visualizations, with only five minutes of active engagement time, plus the necessary time to fully review and audit the results (see my 10-minute video demo here: youtu.be/ZAM9OA0AlUs). To that crucial end of facilitating expert human validation, all projects come complete with a fully reproducible, documented analytic code pipeline and consolidated analytic notebooks for exploration. Then: request revisions, rethink measures, conduct new sub-analyses, run robustness checks, and even add additional deliverables like interactive dashboards, policymaker-focused briefs, and more -- all with just a quick ask to Claude. And all of this can be done *in parallel* with multiple projects simultaneously. By open-sourcing DAAF under the GNU LGPLv3 license as a forever-free and open and extensible framework, I hope to provide a foundational resource that the entire community of researchers and data scientists can use, benefit from, learn from, and extend via critical conversations and collaboration together. By pairing DAAF with an intensive array of educational materials, tutorials, blog deep-dives, and videos via project documentation and the DAAF Field Guide Substack (daafguide.substack.com – MUCH more to come!), I also hope to rapidly accelerate the readiness of the scientific community to genuinely and critically engage with AI disruption and transformation writ large. I don't want to oversell it: DAAF is far from perfect (much more on that below!). But it is already extremely useful, and my intention is that this is the worst that DAAF will ever be from now on given the rapid pace of AI progress and (hopefully) community contributions from here. More than anything, I just hope all of this work can somehow be useful for my many peers and colleagues trying to "catch up" to this rapidly developing (and extremely scary) frontier. It's a wild time, but we need as many people informed and at the table together as possible. Never used Claude Code? No idea where you'd even start? My full installation guide walks you through every step -- but hopefully this video shows how quick a full DAAF installation can be from start-to-finish. Just 3mins! youtube.com/watch?v=jqkVLX… Learn more about my vision for DAAF, what makes DAAF different from other attempts to create LLM research assistants, what DAAF currently can and cannot do as of today, how you can get involved, and how you can get started with DAAF yourself! #vision--purpose" target="_blank" rel="nofollow noopener">github.com/DAAF-Contribut… So there it is. I am absolutely as surprised and concerned as you are, believe me. With all that in mind, I would *love* to hear what you think, what your questions are, what you’re seeing if you try testing it out, and absolutely every single critical thought you’re willing to share, so we can learn on this frontier together. Thanks for reading and engaging earnestly!

Really good advice! Related to your point about testing. I found that one key thing is to remind Claude not to rely on synthetic data. Too often, Claude will try to pass tests by creating idealistic synthetic data. I placed a couple of statements like "- Never create dummy data to fill gaps or pass tests - If data is missing or corrupted, report this honestly - do not fabricate " in my markdown files. Really helps sharpen it up and prevents Claude from taking the easy way.

@kanair I am a big fan of the LDS (and its variants: sLDS, rsLDS) work coming out of the Linderman Lab. Essentially a latent-variable linear model with possion emissions for predicting spike trains. papers.nips.cc/paper/2020/has…

Is anyone building a good predictive model for spike trains?

NWB just turned 10 years old! Researchers worldwide have downloaded 1.9 PB of NWB data from the @DANDIArchive. This animation shows the reach of NWB, facilitating collaboration across the globe. What impact has open neurophysiology data had on your science? Share your stories! 🧠

@bozhi_wu Looks great! What gui framework are you using?

It's always fun to write some useful GUI for your experiment. I started writing this code for controlling the 2p single-cell optogenetic stimulation with TTL. And now I'm just adding more and more useful features into it! #2photon #neuroscience #python #gui #ttl #optogenetics

#eNeuro | Calcium Dynamics in Hypothalamic Paraventricular Oxytocin Neurons and Astrocytes Associated with Social and Stress Stimuli vist.ly/3n59iuu

@bsauerbrei1 Wow this is great! Did you make this yourself? Python?

Finally starting to appreciate the ability of simple models to make prose explanations clearer and more concrete. Here, a model of cortical control of gait modification via phase-dependent gating of sensory input.

@NeuroWoodworks Very pretty! It kinda reminds me of Bull kelp

Manual carving of a 3D neuron in a piece of Walnut. #sciart

The Hypothalamus, a phylogenetically ancient brain region where action potentials and hormones interface to produce hybrid signalling cascades; ultimately enabling our responses to stress. Check Hiro's paper. Amazing work led by @NeuroStress_Lab. biorxiv.org/content/10.110…

Check out our new perspective paper on metastability in neuroscience published @PhysRevE doi.org/10.1103/PhysRe…

@maxime_beau Amazing work! I look forward to trying it out!

I am very proud to announce that my PhD paper finally came out in Cell! In this *very* collaborative study, we develop and release a deep-learning approach to predict neuron type identity from their electrical signature. doi.org/10.1016/j.cell… 1/16 🧵

Last week, I may have DEI’d too close to the sun and attracted anti-NIH folks to my profile who then claimed that my research is useless and a waste of time and money. I beg to differ! Allow me to explain in lay terms how my research in mice is relevant to human health: 🧵

🚨 Exciting news! @NeuroStress_Lab & @Timothy_Bussey have received @CIHR_IRSC funding to explore the role of the prefrontal cortex in behavioral selection under stress. This research could unlock new insights into how stress impacts decision-making! 🧠💡#Neuroscience #Stress

Excited to share my latest work with @DiedrichsenJorn and @andpru. In this work, we ask whether motor sequence learning is motoric at all! Check out the 🧵version of the abstract: biorxiv.org/content/10.110… 1/n

Evolution of the visual system has permeated the basic processing unit in the neocortex, the single neuron. A team effort ⁦@nxwm_network⁩. Single neuron diversity supports area functional specialization along the visual cortical pathways | bioRxiv biorxiv.org/content/10.110…

Congrats @smestern !! Awesome to see @NeuroCarleton alumni do great things!!

@Brains_CAN @WesternuWIN

If you have any questions, want to impletment it yourself, or just want to talk about cool patch-clamp data & open science feel free to contact me!

I’m really excited to share this open-science side project I’ve been working on related to patch-clamp electrophysiology! If you’re into this or have some data you’d like to experiment with, I’d love for you to check it out. Big thanks to Allen Institute for the opportunity!