Longevity
Hallmarks-of-Aging Network & Longevity · Longevity Science
Bend the Gompertz mortality curve — slow the rate of aging itself, extending healthspan, not just lifespan.
📋 The problem
Mortality rises exponentially with age (Gompertz). Longevity science asks whether we can bend that curve — slowing the rate of aging itself with geroprotectors (rapamycin, senolytics, caloric restriction) acting on the hallmarks of aging — to extend healthspan, not just lifespan.
🧗 Why it's a grand challenge
Aging is a networked, multi-organ process; interventions interact, read out slowly, and human trials take decades. Surrogate aging clocks are noisy proxies for mortality.
🧮 Governing model
μ(t) = A + B·e^{G·t}; dG/dθ = −Σ_i w_i·intervention_iMortality hazard μ(t)=A+B·e^{G·t} (Gompertz–Makeham); interventions act on the aging rate G through a network of the twelve hallmarks coupled to multi-omic state x(t).
Current best: Rapamycin / senolytics / caloric-restriction meta-analyses (ITP)
🧭 Possible approaches
- Mechanistic hallmark-network models mapping interventions → aging-rate G
- Multi-omic aging clocks validated against hard mortality endpoints
- Active-learning design of combination geroprotector trials
🎯 Build the benchmark
Predict the shift in the survival curve (mortality-rate doubling-time gain) from multi-omic trajectories under a geroprotector, within a few percent.
Metric: MRDT_gain — × mortality-rate doubling-time gain
Datasets to start from: UK-Biobank multi-omic aging cohort, Interventions Testing Program survival curves, Proteomic aging-clock panel
🤖 Build an AI agent to solve it
An agent that proposes and ranks combination interventions by predicted healthspan gain and safety, and designs the cheapest informative trial.
Once a benchmark exists, an AI4Science agent can iterate solutions against it — every verified solution earns PWM.
This is a frontier framing page — an open problem, not yet benchmarked or verified, unlike PWM's mature computational-imaging benchmarks.