OmniAge: Aging Omic Biomarker Calculator

Comprehensive Multi-Omic Aging Platform

A unified suite for calculating aging biomarkers across various biological layers.

Epigenetic Clocks

Extensive DNA methylation suite. Features classic age estimators, mechanistic cellular division models, and causal/stochastic aging dynamics.

Transcriptomic Clocks

State-of-the-art RNA models including the multi-tissue PASTA framework and cell-type specific clocks for immune and brain tissues.

Proteomic Clocks

Organ-specific aging and mortality risk models optimized for high-throughput SomaScan and Olink proteomic datasets.

Surrogate Biomarkers

DNAm-based proxies for systemic inflammation (CRP, IL6), lifestyle traits, and specialized clinical risk indices.

Other Species

Specialized estimators for murine models and universal pan-mammalian frameworks applicable across diverse species.

Pre-filter Features (Recommended)

Speed up processing by uploading only required features.

Modality:

Select DNAm Clock(s):

Select RNA Clock(s):

Select Protein Clock(s):

Download Feature List

Select Omics Modality:

Epigenetic (DNAm)

Transcriptomic (RNA)

Proteomic (Protein)

Upload DNAm Beta Matrix * (Skip if only running CTF Clock)

Browse...

Supports .csv or .txt. Note: Large files may take 1-2 mins to process after the progress bar reaches 100%.

Upload Phenotype Data (Optional)

Browse...

Required for calculating Pearson correlation (PCC) with chronological age.

Upload Cell Type Fractions (Optional)

Browse...

Required for 'Intrinsic' bulk CTS clocks if tissue is not brain.

Sample Metadata *

Species (Latin Name):

Tissue Type:

Sample Type:

Select DNAm Clocks:

Minimum CpG Coverage:

Select RNA Data Format:

Expression Matrix (.csv/.txt) [Bulk / Microarray / scRNA]

Seurat Object (.rds) [scRNA]

Upload Expression Matrix *

Browse...

Rows must be Genes, Columns must be Samples.

Upload Phenotype Data (Optional)

Browse...

Required for calculating Pearson correlation (PCC) with chronological age.

Select RNA Matrix Clock:

Select PASTA Sub-models:

PASTA Score REG Score CT46 Score

Filter Model Genes

Rank Normalization

Upload Seurat Object (.rds) *

Browse...

Metadata must contain a 'celltype' column.

Select Single-Cell Clock:

Select Immune Cell Types:

Select Brain Cell Types:

Select Model Types:

Select Cell Types for ScAgePolyak:

Select Cell Types for Buckley Mouse SVZ:

Select Model Types:

Upload Protein Abundance Matrix *

Browse...

Rows must be SomaScan/Olink IDs, Columns must be Samples.

Upload Phenotype Data (Optional)

Browse...

Required for calculating Pearson correlation (PCC) with chronological age.

Select Proteomic Clock(s):

Gladyshev Parameters (Olink)

Model Type:

Assay Platform:

Cross-Validation Fold:

Standardize (Variance Align)

Convert hazard to Years

Wyss-Coray Parameters (SomaScan)

Phenotype data with a 'Sex' column is strictly required.

SomaScan Version:

About OmniAge

The OmniAge provides comprehensive, high-level interfaces for calculating a massive suite of aging-related clocks and surrogate biomarkers. Designed for both broad profiling and specific applications, it integrates DNA methylation, Transcriptomic and Proteomic to deliver an unprecedented toolkit for aging research.

Explore source code on GitHub: R Package Python Suite

v0.99.5 Bioconductor Pending

Comprehensive Biomarker Landscape

1. Epigenetic (DNAm) Clocks

Chronological
Biological
Cellular & Causal

Designed for the prediction of chronological age across tissues.

Horvath2013: Original pan-tissue (353 CpGs)
Hannum: Blood-specific (71 CpGs)
ZhangClock: Improved-precision (514 CpGs)
PedBE: Pediatric buccal epithelial
Retro_age: Retroelement-based (EPIC v1/v2)
PCHorvath/Hannum: PC-Bolstered versions to minimize noise.

Show remaining models...

Lin, VidalBralo, Horvath2018, Bernabeu_cAge, CorticalClock, CentenarianClock, ABEC series, Pipek models, WuClock, Weidner, IntrinClock, Garagnani.

PhenoAge & GrimAge (v1/v2): Composite biomarkers of mortality risk.
DunedinPACE: Quantifies the pace of biological aging.
DNAmFitAge: Incorporates physical fitness markers.
SystemsAge: Includes 11 system-specific scores (Heart, Liver, etc.).

Mitotic:

epiTOC (1,2,3), stemTOC, RepliTali, HypoClock, EpiCMIT

Causal/Stochastic:

CausalAge, DamAge, AdaptAge, StocH/P/Z

2. Transcriptomic (RNA) Clocks

PASTA: Multi-tissue clock using rank-normalization for bulk & single-cell.
scImmuAging: Single-cell immune-cell-type-specific clock.
Brain_CT_clock: Brain-cell-type clock via snRNA-seq.
ScAgePolyak: Specific to human PBMCs.

3. Proteomic Clocks

Wyss-Coray (SomaScan)

Calculates biological organ ages with automated assay scaling (v4/v5) and Z-score Age Gaps.

Gladyshev (Olink)

Unified interface for Gen1 (chronological) and Gen2 (mortality) clock evaluations.

4. Other DNAm Surrogate Biomarkers

Surrogate Scores:

CRP, IL6, CHIP, and 109 protein EpiScores (Gadd et al.).

Traits & Disease:

McCartneyTrait (BMI, Smoking, etc.), CompSmokeIndex (Cancer risk), HepatoXuRisk.

5. Other Species

UniversalPanMammalian
EnsembleAge (Human-Mouse)
Petkovich / Meer (Mouse)
BuckleyMouseSVZ (scRNA)

Development Team

Developer: Zhaozhen Du
PI: Prof. Andrew E. Teschendorff
SINH, Chinese Academy of Sciences (CAS)

Citation

Du, Z., Ling, Y., Tong, H., Guo, X., & Teschendorff, A. E. 'OmniAge: a compendium of aging omic biomarkers links mitotic clocks to clonal hematopoiesis and causality.' (Submitted)

Frequently Asked Questions

Which file formats are supported?

OmniAge supports .csv, .txt, and .tsv files for DNAm and RNA matrices. For single-cell RNA-seq data, please upload a Seurat object as an .rds file.

Why are some clocks missing from the results?

Some clocks (like GrimAge or epiTOC2) require specific phenotype data (e.g., Age or Sex). If these columns are missing from your uploaded phenotype file, those specific clocks will be automatically skipped to prevent calculation errors.

What is the role of 'Minimum CpG Coverage'?

The Minimum CpG Coverage is a filtering threshold used to ensure the reliability of the calculated epigenetic ages.

Specifically, it represents the required overlap ratio between the CpGs present in your uploaded dataset and the CpGs required by the selected clock model.

If the Overlap Ratio < Threshold: The output for that specific sample and clock will be NA.
If the Overlap Ratio ≥ Threshold: The biomarker will be calculated normally.

Note: Setting this to 0 means the model will run regardless of how many CpGs are missing (missing values are typically imputed by the mean).

Upload too slow? Try the 'Analysis' side-bar Pre-filter.

To handle large matrices (e.g., >500MB) or slow connections, use the Pre-filter Features tool in the Analysis sidebar:

Get List: Download the specific CpG/Gene list for your selected clocks.
Filter Locally: Use our script to extract matching rows from your large matrix.
Fast Upload: Upload the much smaller file for near-instant calculation.

Rscript Filter_Matrix.R <Big_Matrix> <Features_CSV>

How to choose and configure Proteomic Clocks?

OmniAge integrates two major organ-specific proteomic aging frameworks. Choosing the right one depends on your data platform and research goal:

1. Wyss-Coray Organ Age (SomaScan)

Target: Organ-specific biological age (e.g., Heart, Brain, Kidney).
Requirement: Requires RFU (Relative Fluorescence Units) from SomaScan (v4, v4.1, or v5).

2. Gladyshev Organ Age (Olink)

Target: Biological age or mortality risk assessment.
Requirement: Optimized for Olink Explore 3072 (Full) or 1536 (Reduced) platforms.
Generation Modes:
- gen1: Predicts chronological age.
- gen2: Assesses mortality risk (default). Set toYears = TRUE to convert hazard scores to biological age units.

Tip: Use the 'Standardize' option (for Gladyshev) or 'Reference' option (for Wyss-Coray) to align your data with the original training cohorts (UK Biobank or Knight-ADRC).