Peptides, Astronauts, and Algorithms: Lessons from 10,000 Longevity Experiments
Matt Kaeberlein sits down with the Ora Biomedical team to unpack what their first roughly 10,000 longevity experiments are revealing. They cover a fresh screen of 300 peptides, why most biological pathways look flat until you find the rare high-impact nodes, and how the lab pivoted toward stress resilience, building toward skin-repair products and radiation protection for cancer patients, pilots and future astronauts.
Overview
In this conversation, Matt Kaeberlein talks with the Ora Biomedical team about the lessons emerging from the Million Molecule Challenge, which is now approaching the 10,000-intervention mark in the worm model C. elegans. They walk through a newly completed screen of about 300 peptides tested at two doses with multiple replicates, selected from a mix of known medical compounds, computationally predicted candidates and structurally diverse molecules. The biggest lifespan extension they saw from a peptide was around 20 percent, modest next to the best small molecules, underscoring how hard peptides are to work with because of bioavailability and cell entry.
A central insight is that when you map interventions by pathway, most pathways sit near zero median effect, yet a handful of nodes such as mTOR and IGF-1 can swing lifespan dramatically up or down depending on how they are targeted. The team argues this exposes how thin and flawed much of the training data behind computational drug discovery really is, and why brute-force screening is still needed to generate better data. The discussion then turns to a strategic pivot toward stress resilience, reframing longevity therapeutics as tools that help the body repair from damage.
Their first focus is UV radiation and skin, backed by Space Force funding and a fresh NIH award to build radiation into their robotic platform. They share early results including a mouse study where their best longevity compound cut radiation-induced skin inflammation by about 40 percent, and a natural product that showed early signals in human skin after sunburn. Throughout, they stress a repair-not-blocking mechanism, careful regulatory language, and a topical-first path to market.
Key quotes
5Biggest percent extension of lifespan we saw was about 20 percent. It doesn't compare necessarily to some of our best small molecules which can extend 50, 60 percent.
It's made me so aware of how flawed so much of the training data is out there that people are staking their claims around a discovered molecule on.
I like to joke that we could get to Mars today from an engineering perspective, but we would get there dead.
Our best longevity-extending compound reduced radiation burn dermatitis burden by 40 percent in a mouse model.
We are mechanistically focused on repair from damage, not buffering against damage.
Key ideas
9A 300-peptide screen done the rigorous way
The team screened about 300 peptides at two doses each with multiple replicates, a deliberate contrast to single data points published without context. The dataset, funded by a science-minded philanthropist family, will be released openly.
How the peptides were chosen
Candidates spanned established medical and biohacker peptides, computationally predicted hits aimed at specific pathways, predicted lifespan-shorteners to map the bottom of the curve, and structurally diverse molecules to enrich future screening data.
Peptides are promising but hard to work with
The best peptide extended worm lifespan about 20 percent, below top small molecules. Bioavailability and getting into cells remain hurdles, pointing toward modifications and circularized structures as future directions.
Approaching 10,000 interventions
Ora's database is nearing the 10,000-intervention milestone, roughly ten times the C. elegans data in a major repository like DrugAge. Each added order of magnitude promises more than a tenfold gain in knowledge.
Most pathways are flat, a few nodes are pivotal
Mapping interventions by pathway shows most center on zero median lifespan effect. A few nodes such as mTOR and IGF-1 can move lifespan sharply up or down depending on how they are targeted.
The training-data problem
Much of the data underpinning computational molecule discovery is thin in depth, quality and scope. Brute-force screening is still needed to find the next compounds and to generate better training data for smarter future work.
Reframing longevity as stress resilience
Rather than waiting decades to prove lifespan extension in people, the team frames longevity therapeutics as tools that speed recovery from damage, giving clinical trials a realistic weeks-to-months resilience endpoint.
From worms to mice to human skin
A longevity compound cut radiation-induced dermatitis by about 40 percent in mice, and a natural product showed early human-skin signals after sunburn, compressing years off a typical development timeline.
An mTOR hit and a natural product, topical first
Two lead candidates emerged, one an mTOR inhibitor that is not rapamycin and one a natural-product combination. The team favors topical delivery first to constrain scope and target radiation-induced dermatitis.
Practical takeaways
6- 1
Demand context, not single data points 11:00
Replicated, dose-ranging data beats isolated results. When evaluating any longevity claim, ask whether the finding was repeated and tested across doses.
- 2
Where you target matters more than how much 17:00
The same pathway can help or harm depending on the exact node and direction. Specificity, not just presence of a target, drives the effect.
- 3
Support recovery, not just lifespan 23:30
Repair capacity fades with age. Helping the body bounce back from stress is a tangible, near-term way to think about healthy aging.
- 4
UV is a daily stressor worth respecting 27:00
Sun exposure drives skin aging, sunburn and cancer risk. Both avoiding damage and supporting repair after it have a role in skin health.
- 5
Repair after damage is not the same as blocking it 40:00
A topical applied after sun exposure works on a different mechanism than a sunblock. Don't expect one to substitute for the other.
- 6
Community-backed open science accelerates discovery 42:00
Public donations and open data releases are helping push longevity screening far beyond what any single lab could fund alone.
Topics & chapters
15Back from the Shanghai conference
Ben returns from a longevity conference in China and the team catches up on what moved forward while he was away.
Million Molecule Challenge status
The flagship initiative is still orders of magnitude from a million molecules, but there is exciting new data to share.
The 300-peptide screen
A completed screen of about 300 peptides at two doses with replicates, funded by a philanthropist family and bound for open access.
How the peptides were chosen
A deliberate mix of known compounds, computationally predicted hits, predicted lifespan-shorteners and structurally diverse molecules.
What the peptide results revealed
A top extension of about 20 percent highlights the bioavailability and cell-entry challenges that make peptides hard to work with.
Approaching 10,000 interventions
The database nears a milestone roughly ten times the C. elegans data in a major public repository.
Mapping the longevity network
Most pathways sit near zero median effect, while rare nodes like mTOR and IGF-1 can swing lifespan dramatically.
Why so much training data is flawed
Thin, uneven data undermines computational discovery, so brute-force screening is still essential to build a better foundation.
Longevity reframed as stress resilience
A pivot toward helping the body repair from damage, giving clinical trials a realistic resilience endpoint.
UV damage and skin protection
UV is chosen as the first stress indication, with compounds validated in worms and human skin cells.
Astronauts, pilots and the road to Mars
Early funding from the Space Force frames radiation protection for high-altitude flight and space travel.
The radiation-dermatitis mouse study
A longevity compound reduces radiation-induced skin inflammation by about 40 percent in mice, pointing to cancer-patient benefit.
The NIH RADbot award
A new SBIR grant will build multiple radiation types into the robotic platform to untangle resilience pathways.
The mTOR hit and a natural product
Two leads emerge, an mTOR inhibitor that is not rapamycin and a natural-product combination, with a topical-first strategy.
Repair, not sunblock: early human data
A natural product from traditional medicine shows early signals in sunburned human skin, plus collagen and elastin benefits.