FrontierInvestorLab @Frontier_Lab_

Christian Angermayer has cemented his place in psychedelics by selling biotech AtaiBeckley to Eli Lilly in the biggest acquisition in the industry. "The future is psychedelic, and it is here," he says. Full story: forbes.com/sites/willyako… 📸: Levon Biss for Forbes

A young girl died after receiving an experimental gene-editing treatment in a clinical trial in Shanghai, Science reported, raising questions about regulatory oversight and transparency bloomberg.com/news/articles/…

World models will eat robotics. @bfl_ai x @mimicrobotics

When it comes to supplements to stave off the effects of ageing, columnist Graham Lawton admits he is a hardened sceptic. But he finds some evidence for a few that can improve biological age as we grow older newscientist.com/article/258111…

Aging tissues become stiff. Heart, kidney, lung, and ovarian tissue all accumulate excess structural scaffolding that traps cells and blocks normal function. The process has been considered irreversible, part of the mechanical deterioration that defines aging. A new study tested whether this could be reversed in aging ovaries. Researchers identified the inflammatory signal driving tissue stiffening and blocked it in mice and rats, restoring fertility that had already declined. The findings matter beyond reproduction because the same inflammatory signal drives stiffening across multiple aging organs. That signal is interleukin-11, or IL-11, an inflammatory protein elevated across aging tissues. In ovaries, IL-11 activates fibroblasts to secrete extracellular matrix, the structural scaffolding surrounding cells. IL-11 levels rise in aging heart, kidney, lung, and liver tissue. All of these organs accumulate excessive extracellular matrix during aging. All of them stiffen in ways that impair function. This study tested whether blocking IL-11 could reverse tissue stiffening in ovaries. The ovary served as the model system because age-related stiffness directly blocks follicle development, creating a clear functional readout. Extracellular matrix isn't inherently problematic. Follicles need some matrix structure to develop properly. But when fibroblasts deposit too much matrix, the tissue becomes stiff. That stiffness creates a mechanical barrier that traps follicles and prevents them from growing, maturing, and ovulating. Researchers used atomic force microscopy to measure ovarian tissue stiffness in human samples across different age groups and pathological conditions. The pattern was clear: ovaries nearly tripled in stiffness from the late twenties to early fifties. Here's what they found: • Young ovaries (18-28 years): 1.2 kPa average stiffness, AMH levels 3.9 ng/mL • Middle-aged ovaries (35-42 years): 2.1 kPa, AMH 1.8 ng/mL • Older ovaries (47-52 years): 3.4 kPa, AMH 0.6 ng/mL The kPa measurements represent tissue stiffness, similar to how we measure blood pressure. AMH is anti-Müllerian hormone, the standard clinical marker for how many eggs remain. As ovaries became stiffer with age, the remaining egg supply dropped in parallel. The stiffness pattern showed up in disease states too, not just normal aging. Women with PCOS, premature ovarian insufficiency from chemotherapy, or endometriosis all showed ovarian stiffness levels matching women decades older: • PCOS ovaries: 3.2 kPa (comparable to 47-52 year olds) • Chemotherapy-induced POI: 2.8 kPa • Ovarian endometriosis: 2.9 kPa • Age-matched healthy controls: 1.3 kPa These conditions accelerate the mechanical aging process, producing the same tissue stiffness that normally takes decades to develop. POI is premature ovarian insufficiency. All three conditions showed tissue stiffness comparable to ovaries decades older than the chronological age of the patients. To identify what drives this stiffening, the team analyzed proteins in human ovarian tissue and gene expression in human ovarian fibroblasts. IL-11 emerged as a central regulator. In aging ovaries, IL-11 expression was significantly elevated compared to younger tissue. IL-11 works through a straightforward pathway. The protein binds to a receptor on fibroblast cells, which activates internal signaling that tells those cells to produce more collagen and other structural proteins. More structural proteins means stiffer tissue. The researchers tested whether blocking this pathway could prevent the stiffening. They used three different methods in mice and rats: removing the IL-11 receptor genetically, using antibodies to block the receptor, and using RNA interference to silence IL-11 production directly. All three approaches worked. In middle-aged mice with the IL-11 receptor removed, ovarian tissue stayed soft. At 10 months old (middle age for mice), these animals had tissue stiffness of 1.8 kPa compared to 3.2 kPa in normal aging mice. That's nearly half the stiffness of their age-matched counterparts. Functional outcomes improved alongside the mechanical changes: • IL-11 receptor knockout mice maintained higher follicle counts across all developmental stages (big deal) • Primordial follicle reserve was preserved longer (another big deal) • Ovulation rates remained higher in middle-aged knockout mice • Fertility extended into later reproductive age The team tested the same intervention in disease models. In chemotherapy-induced POI, blocking IL-11 signaling prevented the stiffness increase that normally follows chemotherapy treatment. Follicle counts and hormone levels remained closer to baseline compared to untreated controls. PCOS affects about 10% of women of reproductive age. One of the defining features is irregular ovulation, where follicles either don't release eggs or do so unpredictably. The standard explanation focuses on hormonal imbalances, but this study suggests the mechanical environment plays a role too. In PCOS mice, blocking IL-11 reduced ovarian stiffness and restored more regular ovulation patterns. The intervention worked at the level of tissue mechanics, not just hormone levels. When the tissue became less stiff, follicles could develop and release eggs more consistently. Single-nucleus RNA sequencing revealed how blocking IL-11 changes the cellular composition of ovarian tissue. In aging ovaries, a specific population of fibroblasts becomes hyperactive and expands over time. These cells produce excessive amounts of structural proteins and are locked in a state of continuous activation. The activation is driven by the ERK pathway, one of the key signaling systems cells use to respond to growth signals and stress. When ERK is constantly active, it keeps telling fibroblasts to produce more matrix proteins. This creates the stiffening feedback loop that defines ovarian aging. When IL-11 signaling was blocked, the proportion of activated fibroblasts decreased significantly. The fibroblast population shifted back toward a less activated state with lower matrix production. This explains the tissue-level stiffness reduction at the cellular level. The most clinically relevant finding came from the therapeutic intervention experiments. Researchers treated already-aged mice and rats with nanoparticles carrying siRNA targeting IL-11. These nanoparticles were designed to accumulate specifically in ovarian tissue after systemic administration. In 10-month-old mice treated with IL-11 siRNA nanoparticles for 8 weeks: • Ovarian stiffness decreased from 3.2 kPa to 2.1 kPa • Follicle counts increased across developmental stages • Estrous cycles became more regular • Fertility rates improved compared to age-matched untreated mice The same treatment worked in aging rats. 12-month-old rats receiving IL-11 siRNA showed reduced ovarian stiffness, improved hormone profiles, and extended fertility compared to controls. This matters because it demonstrates reversibility. The tissue stiffening that accumulates during aging isn't permanent structural damage. It's an active process maintained by ongoing IL-11 signaling. When that signal is removed, even in already-aged tissue, the mechanical environment can be restored. The mechanism fits into the broader understanding of ovarian aging. Chronic inflammation increases with age in most tissues, a process called inflammaging. IL-11 is one component of that inflammatory environment. Previous research has focused on IL-6 and TNF as drivers of age-related inflammation, but this study identifies IL-11 as specifically relevant to the mechanical aspects of ovarian aging. The study provides mechanistic detail on how this happens. In follicles cultured in stiff environments, genes related to follicular growth were downregulated while inflammation and ECM reorganization genes were upregulated. The mechanical signal creates a feedback loop where stiffness promotes more inflammation and more matrix deposition. There are some really critical limitations of this mouse model study. The human tissue measurements are correlational. Higher stiffness associated with aging and disease, but causality was established only in animal models. Whether blocking IL-11 in humans would produce similar functional improvements requires clinical testing. The interventions were preventive or early therapeutic in the animal models. Mice and rats were treated before severe ovarian dysfunction developed or shortly after aging began. Whether the same approach works in late-stage ovarian insufficiency or after menopause isn't addressed by this data. Safety and off-target effects need evaluation. IL-11 has roles in other tissues, including wound healing and inflammatory responses. Long-term suppression could have consequences beyond the ovary. The nanoparticle delivery system aims to concentrate the intervention in ovarian tissue, but complete specificity is unlikely. Still, the therapeutic concept is supported by converging evidence. Surgical interventions that reduce ovarian stiffness, like ovarian fragmentation or drilling, have shown clinical efficacy in restoring fertility in some cases of POI and PCOS. Those procedures work mechanically. An IL-11 antagonist would achieve the same mechanical outcome through a molecular intervention. Clinical translation faces practical hurdles. The optimal dosing, delivery route, and treatment duration for IL-11 antagonism in human fertility applications haven't been determined. Whether the treatment works for all causes of ovarian dysfunction or only those driven primarily by matrix stiffening is unknown. What the study establishes clearly is that IL-11 is a druggable target that regulates tissue mechanics across aging. Matrix stiffness isn't a passive consequence of getting older, it's an active inflammatory process that can be modulated. Blocking one inflammatory protein is sufficient to preserve tissue mechanics and extend functional outcomes in multiple animal models. The ovarian findings matter because they prove a concept that applies across tissues. IL-11 is elevated in aging heart, kidney, lung, and liver tissue. All of these organs accumulate excessive extracellular matrix during aging. All of them stiffen in ways that impair function. If the same IL-11 pathway drives mechanical aging across tissues, this represents a single intervention point that could address tissue stiffening systemically. The convergence across aging, PCOS, POI, and endometriosis in ovaries suggests IL-11-driven stiffening may be a common pathway in multiple forms of tissue dysfunction. If confirmed in humans, this represents a mechanistic intervention that addresses a shared feature of age-related organ decline rather than treating each tissue separately. Right now, we know IL-11 causes tissue stiffening that blocks normal function, and blocking it restores mechanical and functional outcomes in animal models. The ovary served as the proof of concept. Whether that translates to reversing tissue stiffness in other aging organs is the next question.

I nearly shut down my longevity company last year. I was asking myself what one should build on the eve of superintelligence. The answer is obvious now: cure all diseases. But this wasn’t an obvious path for an individual like myself 12 months ago. Sid Sijbrandij hadn’t yet put his cancer into remission by going founder mode. The Australian engineer hadn’t shrunk his dog’s tumor by building a custom mRNA vaccine. And Kate and I hadn’t yet been diagnosed with disease. Curing disease was something that big, billion dollar companies did over ten years. Not individuals. That’s now changed. Two weeks ago, I started focusing my company, Immortals, on building the infrastructure to allow individuals to discover and resolve their own health issues. Kate and I are the first customers as we both try to address our recent disease diagnoses. If you’re a cracked engineer looking for a hard challenge, come build this with me. Immortals will continue to be best-in-class with nutrition, GLP-1s, longevity medicines, hormones, peptides, biomarkers and concierge medicine. We’re now adding induced pluripotent stem cells, organoids, deep cellular characterization, and personalized therapy development among other things. I hope what we build in the coming year makes current health care look like it’s medicine from the 1500s. Hit me up if you’re building biotech that matches in goal. When our basic needs are met, solving all diseases is the only rational next thing to do. And to not wait around for someone else to do it on your behalf.

Bryan Johnson @bryan_johnson

a year ago

Blueprint has been a pain in my ass. It's kept me from not focusing on the single thing I’m consumed with: how does the human race survive the rise of super intelligence. Every minute spent dealing with problems like ‘why a supplier shipped us something out-of-spec’ (now

824 342 10K 1.1M 2K

x.com/i/article/2080…

Aging doesn't happen in a straight line. The biggest molecular shift in the brain occurs around midlife, suggesting aging may accelerate in stages rather than gradually. Why this happens is not known

Biology is stuck in a tradeoff that semiconductors escaped decades ago: we constantly "dumb down" data richness (like using flat petri dishes) just to "amp up" our throughput (getting more rows of data). But semis broke this exact SAME bottleneck! It is not an unfamiliar problem. Integrated circuits took a complicated process and scaled it, allowing complexity and throughput to compound together. For the last two years, our team has been building biology’s equivalent. Right now, the world is obsessed with AI drug design. But design is 20% of the problem [Dr. Lowe, Novartis] and it IS commoditizing. The bottleneck is now accepted as having shifted to the 80%. AI can’t help here yet because we lack the data richness to train it. Pharma has been trying to do calculus with data that is human-native; AI labs try to make-do with what data they get from a vendor and invest their money into compute and models. But there is no database an AI can exponentiate its understanding of human biological response from. There's fragments of data, far too simplistic. A pathology slide hands you a frozen state. But biology isn't ONE state or ONE stain; it's an active process. Drug response is a trajectory - just, too hard to industrialize. Sound like semis? Consumer electronics? Well, the answer we found was the EXACT SAME: a CHIP. And laser biophotonics for cell based sensing. (1/)

Robotics startup Genesis in talks to raise around $500 million at a $3 billion pre-money valuation bloomberg.com/news/articles/…

AlphaFold3 predictions of the molecular contacts made by a genome-editing enzyme have been used to improve the enzyme’s selectivity for target DNA sequences go.nature.com/4wdC4vT

Formation Bio said it has appointed former Biogen research head Michael Ehlers as its chief scientific officer, bringing a seasoned drug and biotech veteran into the AI-focused health startup. bloomberg.com/news/articles/…

Genomics has quietly entered a new phase: AI agents that autonomously discover, configure, execute, and chain bioinformatics operations from plain natural-language instructions. This Cell Genomics perspective names it "agentic genomics" — delegating multi-step genomic analyses to autonomous agents that select tools, manage dependencies, and adapt execution based on intermediate results, all mediated by LLMs and constrained by domain-specific skill libraries. The authors' central claim: agentic genomics doesn't just speed up pipeline construction — it shifts the actual bottleneck in computational biology to validation. They survey emerging systems (CellAtria, AutoBA, Bio-Copilot, ClawBio) and their divergent architectures, then propose a tiered validation framework spanning research-grade, benchmarked, and clinical-grade analyses — arguing equity-aware design has to be a systems requirement, not an afterthought, for agentic genomics to actually be trustworthy. cell.com/cell-genomics/…

Really exciting progress from general-purpose LLMs on ADMET prediction. The accuracy still has room to improve, but the pace of advancement is impressive. Huge credit to @AnthropicAI, @DarioAmodei, and everyone building and improving these systems.

Bogdan A. Zagribelnyy, PhD @sumrexromanus

4 days ago

🤖Claude Opus 4.8 ranks among the top LLMs for drug ADMET prediction. 📈 Claude Opus 4.8's ADMET comeback is incredible: version 4.6 stumbled, finishing #10 out of 13 on prediction error and near the bottom on compound ranking on our #ADMET benchmark sets that we thoroughly

2 7 33 6K 6

With artificial-intelligence tools speeding up and taking on ever more steps in science, we asked our readers which tasks they would rather keep for themselves go.nature.com/4wgqsbK

A mathematical analysis suggests that the notion of agency, which is a prerequisite for consciousness, cannot be purely quantum in nature newscientist.com/article/257954…

Anthropic is launching a new Rare Disease Research Grants program. Selected researchers and early-stage biotech companies can receive up to $50,000 in Claude credits over six months to support mechanism discovery, data analysis, drug development workflows, and regulatory document preparation. Application deadline: August 2, 2026. anthropic.com/news/rare-dise…

Anthropic @AnthropicAI

6 days ago

We're offering grants of up to $50,000 in Claude usage credits to researchers accelerating cures for rare diseases. This is our first focused call within AI for Science, our program supporting scientists using Claude to speed up discovery. anthropic.com/news/rare-dise…

728 1K 8K 1.9M 2K

A lot happened in AI × bio over the past week. Here’s what you might’ve missed: ☑️ Recent AI critic Jennifer Doudna’s lab used AI to rewrite a CRISPR-like gene editor - and it worked like a charm. ☑️ A startup made a system that can repeatedly read the same living cell without killing it. That is huge! But... ☑️ Two years ago Chai barely existed. Now the company has raised $400M at a $3.8B valuation, with Lilly, Pfizer and Novartis on board. ☑️ Lila says it sees "local spikes of superintelligent behavior". Can you quantify that? No, but it sounds cool! 🧵 (1/6)

Paper: nature.com/articles/s4146…

Hello world 👋 We're building the first medicines to repair the aging extracellular matrix, the scaffold that holds your body together. Our new results in Nature Communications got picked up by @theallinpod, @bryan_johnson, @davidasinclair, @afshineemrani, and many others. What did we find? Let's dig in 🧵 Cells get almost all the attention in biology. But cells don't live in a vacuum; they're held up by structural tissue, the extracellular matrix. Collagen, elastin, and the scaffolding that gives your body its shape. As the matrix accumulates chemical damage over time, its properties change. Skin sags, arteries stiffen, and the damaged matrix draws in immune responses and drives oxidative stress. It's one of the root causes of aging, and it's been largely neglected. Natively, our bodies can't repair this damage. But some bacteria have enzymes that can break these chemical bonds. So we asked: can we engineer highly efficient, therapeutic versions of these enzymes? Many said it couldn't be done. We built a platform to do exactly that. CML forms via the same Maillard reaction that browns bread: slowly, over a lifetime, at body temperature. Protein-bound CML has been considered irreversible for 40 years. You can't untoast bread. 🍞 So we built CMLase, an enzyme that clips CML off proteins and restores the original. We screened tens of thousands of natural enzymes, then ran directed evolution across 500 million variants of the best ones. In vitro, it works. More importantly: CMLase reverses CML in real ex-vivo human tissue from elderly donors. This is the test that matters. So many aging interventions look great in mice and fail in humans. After applying CMLase overnight to tissue from elderly donors, residual CML dropped to levels seen in far younger tissue. In elderly skin: a 55%+ reduction, down to below the level of 31-year-old skin. This is just the start. We've shown proof-of-concept that engineering enzymes to rejuvenate the extracellular matrix is possible: untoasting the proverbial bread. We're raising our Series A soon. If you believe, like @chamath at the @theallinpod, that this is a trillion-dollar market, get in touch.