←
Norn Group507 posts
Norn Group avatar✓

Norn Group ✓

@NornGroup

Norn Group is a do tank for industrializing longevity. We are working towards a world that creates years of healthy life faster than time passes.

◷ Joined November 2022▧ 204 Photos and videos
Norn Group avatar
Norn Group@NornGroup·Nov 16

New @impetusgrants-funded paper on predicting cell-state changes is out in @NatureProtocols🚨🚨🚨 Stanford’s Xiaojie Qiu and MIT's Jonathan Weissman have published a guide to using Dynamo, a computational framework first introduced in Cell in 2022. Dynamo infers a model of cell-state dynamics by integrating single-cell gene-expression data with estimates of the rate of change in expression. It can be used to make predictions about cell trajectories, the effects of perturbing genes, and prioritizing putative regulators of cell fate conversion. The authors present a new protocol describing these analyses with nine case studies on four types of single-cell data, and includes code, notebooks and example datasets. In an example using Dynamo, they showed that in silico activation of KLF1 skews cell fate decisions toward red blood cell lineages, whereas silencing KLF1 and activating GATA1 results in a prediction that cells will remain in a progenitor state that gives rise to platelet-producing cells and red blood cells. Making AI useful for longevity takes both the right biological data and methods that turn it into predictions researchers can test. We’ve been making this case for a while, and funding those datasets will be the focus of our next @ImpetusGrants round: norn.group/impetus-grants/ro… Read the paper: doi.org/10.1038/s41596-026-0…

Post attachmentPost attachmentPost attachment
Norn Group avatar
Norn Group@NornGroup·Nov 14

AI in bio needs more data. But “more” isn’t a strategy. What data does AI need to accelerate longevity drugs? Our founder @MartinBJensen is back in the arena today at Foresight Institute’s Vision Weekend USA in San Francisco, speaking on “Making AI Work for Longevity.” He’ll be talking about how the data we generate determines where and when AI accelerates progress. New cures need physiological data from organs and organisms, and those experiments take months or years. We need to start generating that data now. Event program: visionweekend.foresight.org/

Post attachmentPost attachment
Norn Group avatar
Norn Group@NornGroup·Nov 12

Yes, there’s a chance AI helps us cure these diseases within a decade. But AI can’t get there without the right training data, and some of it takes years to collect. We should start now. @DarioAmodei, are you in? norn.group/impetus/round4

Post attachment
1:04
Norn Group avatar
Norn Group@NornGroup·Nov 5

Agreed! We funded Columbia’s VIBRANT study via @ImpetusGrants because we wanted to move ovarian aging research into human trials. Led by Yousin Suh (@YousinSuh) and Zev Williams, the trial is testing whether low-dose rapamycin can delay menopause by slowing ovarian aging, and what that could mean for women’s reproductive health. More info here: clinicaltrials.gov/study/NCT…

Post attachment
Norn Group avatar
Norn Group@NornGroup·Nov 5

Some cell populations become more common with age. Others become less common. Why is that? Can we use what changes in these populations to understand how tissues age? We funded Junyue Cao @junyue_cao at The Rockefeller University via @impetusgrants to study how programmed cell death and cell populations change across cell types as mice age. His team has given the field two large, publicly accessible datasets, presented in two Science papers. In the first paper, Cao and colleagues measured gene activity in almost 22 million cell profiles across 14 mouse organs and tissues. They found that some cell populations became less common during early adulthood, while several immune cell populations became more common later in life. They also looked at mice without functional B and T cells and compared them with normal mice. The effects differed between cell populations. Certain kidney and lung populations did not expand with age in these mice. Most populations that became less common with age followed the same pattern as in normal mice. In their second paper, Cao and his colleagues measured how accessible different regions of DNA were to gene-regulating machinery in cells from 21 tissues. Some changes were shared across tissues, while others differed between sexes. The relative abundance of one kidney cell type, for example, increased with age in female mice but decreased in males. The DNA-accessibility atlas also revealed that areas associated with regulating inflammatory genes generally became more accessible with age in many cell types. Read the papers: doi.org/10.1126/science.adn3… doi.org/10.1126/science.adw6…

Post attachmentPost attachment
Norn Group avatar
Norn Group@NornGroup·Oct 30

Who tests the assumptions behind aging biomarkers? How do we know they measure aging the same way in different people? We funded Alan Cohen via @impetusgrants to compare markers of immune aging across different populations. Inflammaging is the increase in chronic inflammation that happens as we get older. Previous research in Italy had identified a pattern of inflammatory blood markers that was associated with aging and chronic disease. But how universal was that signature? In Nature Aging, Alan Cohen, Michael Gurven and colleagues compared inflammatory markers in four populations. In Italy and Singapore, levels of inflammatory blood markers were higher among older adults and those with chronic diseases (especially kidney disease). In a Bolivian and a Malaysian population, however, inflammatory markers were not (or only weakly) related to age and the relationships with chronic diseases did not generalize. Even Italy and Singapore differed on individual markers. IL-6, commonly used as a marker of inflammaging, was not associated with age in the Singaporean population. Read the paper: doi.org/10.1038/s43587-025-0…

Post attachment
Martin Borch Jensen avatar
⟳ Norn Group retweeted
Martin Borch Jensen@MartinBJensen·Oct 28

Biology's long timelines are often brushed aside as something that can be brute-forced away with more compute. The arguments goes: (1) robots can speed up experiments, (2) enough compute can simulate biology, and (3) smarter AI can discover new interventions from existing data. I think postponing data collection for future robots/whatever is a mistake. "Can't robots run the experiments faster?" To some extent, yes! Robots can speed up measurements and run experiments in parallel. But observing how organs or diseases change over time, or how interventions change that course, involves incompressible time. Robots can speed up in order to... wait. "Can't we simulate, what about AlphaFold?" It benefited from 50 years (!) of protein-structure data. You can't check your simulations without data. Learning how aging and disease unfold, and what changes their course, requires physiological datasets that can take years to collect, however quickly models improve. The argument isn't to collect data as usual, ignoring or pausing AI. Collecting data alongside training better models puts us on the most optimistic trajectory where AI is maximally useful for aging and disease. We (certainly anyone promoting AI cures) should allocate a fraction of the budget to high-quality data collection.

Post attachment
Norn Group avatar
Norn Group@NornGroup·Oct 28

"Alphafold solved this subset of static structure prediction with a dataset that took 50 years to collect." If we hope AI will help with aging, we should start generating the data it will need now. If we wait until we need them, we’ll be disappointed. It’s not a question of whether we should pause AI development and instead just collect data. We can take a micro fraction of the overall budget and use it to generate the right data in parallel with everything else. The next Impetus Grants @ImpetusGrants round we're setting up is a direct path to that. Our goal is $15M to get these datasets started, and we’ve raised ~$9M so far. norn.group/impetus-grants We already anticipate compute demand and build data centers years before we need them. We should do the same for training and verification data on aging and disease.

Post attachment
Martin Borch Jensen avatar
⟳ Norn Group retweeted
Martin Borch Jensen@MartinBJensen·Oct 26

Re. compute vs bio data collection, path is obvious when you consider EV of marginal spend on either: If a frontier lab put $50M into collecting data that may be gating, e.g. rich multiomics from real human longitudinal trials, that's ~1 day of compute budget. Such data takes incompressible time from study initiation. If pure simulation doesn't need this data, cures delayed 1 day. If the data guides models to simulate sooner, cures accelerated by months or years. If the data is strictly necessary, cures accelerated years. Obvious choice if eg the first AI benefit example in your S-1 is ""profoundly accelerate the rate of progress in biology and medicine, potentially curing diseases that have plagued humanity for millennia ..."

Post attachment
Norn Group avatar
Norn Group@NornGroup·Oct 25

Agree. Some questions (like how a disease develops, or whether a drug has long term health benefits for patients) can only be answered by following people over years. If we don't start those studies now, then future AI will not have access to that data. Lots of things we could do: we can follow patients’ health over time, see how they respond to treatment, and even bank blood samples for tests that may become available later. Laboratory experiments that take days or weeks can run alongside those studies. We’re raising $15M for the next round of @ImpetusGrants to fund work like this, with ~$9M raised so far. More here: norn.group/impetus-grants

Norn Group avatar
Norn Group@NornGroup·Oct 23

If we’re optimistic about what AI will do for medicine, we should be collecting the data it will need to deliver. In other words: we already plan for future compute demand by building datacenters. We should also build biological datasets to meet anticipated demand from abundant intelligence. Our next Impetus round will fund AI-enabling datasets and the tools to generate them. We’ve raised $9M toward our $15M goal. If you’d like to help build the datasets that future biomedical AI will depend on, donate here: norn.group/donate That includes data linking molecular and cellular changes to measured outcomes in organs and people. Some projects will start experiments that take years to mature; others can build on years of observation already completed. Projects could include: - Existing cohorts and trials: adding richer measurements of stored samples and linking them to recorded health outcomes. - Experiments in living organisms: testing interventions and measuring cellular responses within diseased organs. - New measurement tools: extracting more biological information from routine samples, such as blood. - Adaptive clinical trials: testing AI-generated hypotheses and adapting as evidence accumulates. Learn more: norn.group/impetus-grants

Post attachmentPost attachment