epi-clock — Horvath Epigenetic Age Calculator

// interactive calculator

⌨️ Input

Chronological Age (optional)

CpG Site	Gene	β Value

📊 Result

🧬

Load demo data or enter CpG values, then click Calculate

// demo output

$ python3 epi_clock.py demo

╔══════════════════════════════════════════════════════════════╗ ║ HORVATH EPIGENETIC CLOCK — DEMO MODE ║ ╠══════════════════════════════════════════════════════════════╣ ║ Three synthetic samples: ║ ║ 🟢 Young adult (target: 25 years) ║ ║ 🟡 Middle-aged (target: 50 years) ║ ║ 🔴 Elderly (target: 75 years) ║ ╚══════════════════════════════════════════════════════════════╝ SAMPLE: Young adult (25 years) ┌─────────────────────────────────────────────────────────┐ │ 🧬 Predicted Biological Age: 21.2 years │ │ Raw DNAm age (pre-transform): +0.0566 │ └─────────────────────────────────────────────────────────┘ Biological Age........ ████████░░░░░░░░░░░░ 21.2 yrs Chronological Age..... ██████████░░░░░░░░░░ 25.0 yrs Age Acceleration: -3.8 years 🐢 DECELERATED AGING Confidence: ●●●● HIGH SAMPLE: Middle-aged adult (50 years) ┌─────────────────────────────────────────────────────────┐ │ 🧬 Predicted Biological Age: 46.2 years │ └─────────────────────────────────────────────────────────┘ Biological Age........ ██████████████████░░ 46.2 yrs Chronological Age..... ████████████████████ 50.0 yrs Age Acceleration: -3.8 years 🐢 DECELERATED AGING SAMPLE: Elderly (75 years) ┌─────────────────────────────────────────────────────────┐ │ 🧬 Predicted Biological Age: 71.2 years │ └─────────────────────────────────────────────────────────┘ Biological Age........ ████████████████████████████░░ 71.2 yrs Chronological Age..... ██████████████████████████████ 75.0 yrs Age Acceleration: -3.8 years 🐢 DECELERATED AGING

// how it works

📐 The Math

DNAm_age = intercept + Σ(coeff_i × β_i)

age = exp(DNAm_age + log(21)) − 1

Elastic net regression on 353 CpG sites, with a log-linear calibration for developmental ages. The model intercept is 0.6955.

🧬 CpG Sites

This implementation uses the 36 most influential CpG sites (by coefficient magnitude) from the full 353-site Horvath clock. Key markers:

ELOVL2 — strongest age-correlated CpG
FHL2 — robust aging marker
NHLRC1 — largest positive coefficient
RPTOR — mTOR pathway link

⚡ Age Acceleration

ΔAge = biological − chronological

Positive → accelerated aging (biologically older)
Negative → decelerated aging (biologically younger)
Predicts all-cause mortality independently
Used to evaluate anti-aging interventions

📊 Related Clocks

Horvath (2013) — 353 sites, pan-tissue
Hannum (2013) — 71 sites, blood
PhenoAge (2018) — 513 sites, mortality
GrimAge (2019) — 1030 sites, mortality
DunedinPACE (2022) — 173 sites, pace

// cli usage

usage