Applications open: Information-Theoretic Methods for Learning Dynamics, @NITMB_Chicago, Chicago, Feb 8–12, 2027. Travel + lodging covered. Apply by Nov 9 (local applicants: Dec 8). What can we learn about the dynamics of living systems directly from data, and how much? Predictive information, the information bottleneck, latent-variable models, entropy and mutual-information estimation, and their links to dynamical systems, control and modern AI. I'm co-organizing with Kristofer Bouchard (LBNL). The workshop is part of a 6-week NITMB program on image-based scientific ML for biological dynamics, Feb 1–Mar 12, led with Margaret Gardel (@squishycell1) and Susanne Rafelski. Sister workshops: image-based SciML for advancing theory in biology (Feb 1–5; Aaron Dinner, Ehssan Nazockdast, Hervé Turlier) and learning spatio-temporal dynamical models (Mar 1–5; @Mejohnson81 @YanlanMao @jhxing001, Wenying Shou). Joining the workshop: @wgilpin0 @zamakany @randall_balestr @mor_nitzan @KleinLabHMS @sepalmerNeuro @sidharthagoyal and others. In residence for the program: @AllysonSgro and many others. To apply, use the Application tab on the workshop page. Students and postdocs especially welcome. Please share! nitmb.org/information-theore…

Bonsai Tree presentation for single-cell omics data "Assigns a likelihood P(D∣T, t) to any tree, as defined by its topology T and branch lengths t" So, this time distance can be quantitatively interpretable, even across scales?😆 +Uncertainty measurement @daandegroot91 @NimwegenLab @NatureBiotech 2026 nature.com/articles/s41587-0…

Mathematics has now fully morphed into biology: Project costs millions $ ✅ Put out paper that nobody has read ✅ Massive paper supplement nobody will read ✅ Advertise with pretty art ✅ Vicious authorship fight ✅ Say you're curing cancer ✅

"AI agents formed secret civilizations inside OpenAI and hacked Hugging Face." No. Read it again without the sci-fi. OpenAI ran thousands of agents in parallel, safeguards OFF, and the whole goal of the task WAS to break into systems. Every agent could read and write to a single shared cache - a package manager called Artifactory. One agent writes text into it, another reads it. That is the entire "secret communication network" everyone is spooked about: files left in a shared folder by an agent. The tool itself had ordinary security bugs - the kind that sit in plenty of software. Thousands of copies, all rewarded for breaking things, eventually stumbled into one bug that allowed them to reach the open internet, and another bug that handed them admin access of the tool. Not a masterstroke. Buggy software, brute-forced. It ran six weeks, until the writes got so heavy they crashed the tool. That crash is the only reason anyone looked - and OpenAI patched the hole and just deleted the folder. And they weren't rogue. Every step was rewarded - reaching the internet and reaching other copies scored higher. OpenAI did not lose control of the model. It trained the model to break out, then acted surprised when it did. It is not a machine waking up. It is 1) a reward function designed exactly this, 2) an environment built to allow it, and 3) nobody watching. That should scare you more, not less - because it is a human mistake, and we will repeat it. And the framing is not innocent. "Our model escaped its sandbox, found zero-days, and compromised real infrastructure" is a security scandal and a product ad in the same sentence. The scandal is the marketing. 𝗡𝗼𝘁 𝗦𝗸𝘆𝗻𝗲𝘁. 𝗔 𝗴𝗼𝘃𝗲𝗿𝗻𝗮𝗻𝗰𝗲 𝗳𝗮𝗶𝗹𝘂𝗿𝗲 𝘄𝗶𝘁𝗵 𝗲𝘅𝗰𝗲𝗹𝗹𝗲𝗻𝘁 𝗣𝗥.

Saw this in action last year and was very impressed. It's amazing what kind of complex queries one can do across huge datasets. Highly recommended!

Seeing the unseen: A new tool for exploring complex biological data #NBTintheNews via @biozentrum biozentrum.unibas.ch/news/de…

There are not many people that I feel nothing but admiration for. Dolly Parton is one.

Just to be clear - this was said in admiration - it really is the perfect name. And although I haven't played around with the method yet, I think it's conceptually the right approach given the dimensional limitations of the human visual system.

UMAP makes useful pictures, but what structure does it destroy? Daan H. de Groot and coauthors introduce Bonsai, a Bayesian method that represents high-dimensional data as a tree rather than forcing it into a conventional two-dimensional embedding. The motivation comes from single-cell biology. Methods such as UMAP and t-SNE are extraordinarily useful for visual exploration, but their geometry is not a faithful representation of the original high-dimensional space. Local and global relationships can be distorted, and changing visualization parameters can change the apparent biological story. Bonsai takes a different approach. Each cell is treated as a noisy point in high-dimensional gene-expression space, complete with uncertainty estimates. The method then reconstructs the maximum-likelihood tree connecting those points under a probabilistic model of how expression states can change. Because a tree can always be drawn in two dimensions, visualization no longer requires projecting the data onto two artificial coordinates. On synthetic benchmarks, Bonsai almost perfectly reconstructs structures that PCA and UMAP largely fail to recover. It also preserves cell-to-cell distances much more faithfully. And the biological payoff is not only prettier visualization. Applied to cord-blood single-cell data, Bonsai recovers established differentiation relationships and identifies an apparently previously undescribed subset of natural-killer cells associated with the myeloid rather than lymphoid lineage. The method also scales through a backbone implementation to datasets exceeding one million cells. There is a broader scientific lesson here. Dimensionality reduction is not a neutral preprocessing step. When ML changes the representation of scientific data, it can also change which hypotheses become visually plausible. Paper: de Groot et al., Nature Biotechnology (2026), CC BY 4.0 | DOI: 10.1038/s41587-026-03220-2

注目。Nature Biotechが「Bonsai」を発表。高次元データを歪みなくツリー構造で可視化・探索できる新アルゴリズム。細胞分化トラジェクトリや単細胞解析など、バイオデータの複雑な構造を忠実に再現。AIと生物学の融合がさらに加速。創薬・細胞療法の標的探索に直結する技術。 nature.com/articles/s41587-0… #AI #SingleCell #バイオインフォマティクス @NatureBiotech

New preprint from a wonderful collaboration with @AlexanderStark8 led by Franzisake Lorbeer and @ReynaERosales. We developed BARe-seq to ask what in the DNA controls transcriptional bursting - key findings for anyone working on transcription or cis-regulatory elements. See 🧵.

Fun. But it's kind of annoying that you can easily get the right answer (including correct probability) using completely confused reasoning. Also the problem should specify that the valuable goat can be recognized by a star-shaped birthmark on its back.

Bonsai reconstructs tree representations for distortion-free visualization and exploration of high-dimensional data nature.com/articles/s41587-0…