Deep Biomarkers Of Human Aging


How old by basic blood test

Aging.AI1.0

  • 41 input parameters
  • r = 0.91
  • Rsq = 0.82
  • MAE = 5.5 years
Test your samples

Aging.AI2.0

  • 33 input parameters
  • r = 0.79
  • Rsq = 0.63
  • MAE = 6.2 years
Test your samples

Aging.AI3.0

  • 19 input parameters
  • r = 0.80
  • Rsq = 0.65
  • MAE = 5.90 years
Test your samples

Floro'clock

  • 1066 input parameters
  • r = 0.5
  • Rsq = 0.3
  • MAE = 5.9 years
Test your samples


Please try our Young.AI - a tool for tracking your predicted age over time using the multiple data types



This is a deep-learned predictor of your age made with a deep neural network trained on hundreds of thousands anonymised human microbiome profiles.

Floro'clock is based on an improved version of the model described in Galkin et al

Please use the citation when referencing this project

Upload your gut micrflora profile obtained from WGS sequencing here to predict your age. Things to check before uploading:

We recommend to obtain microflora profile with Centrifuge software. Human gut contains thousands of different microbes, but each individual profile ususally contains no more than several hundreds of reliably detected species. It is OK to provide tables with fewer columns than 1066 columns: all the missing values will be filled with zeroes on our end.




You can upload up to 25 samples per session without creating an account. Log in to increase your session limit to 100 samples.

How to partner with us

We are driven by our mission to extend healthy human longevity and when you partner with us, you contribute to solutions that benefit everyone.

If you’re interested in sponsoring a research project or simply accessing our extensive research infrastructure, we’ll help you launch a successful and rewarding collaboration with researchers who are leaders in their fields.

We closely work with academic partners to develop deep biomarkers of human aging and health status.

View partnership case studies.

Check out Insilico Medicine publications

Contact the Pharma.AI research team for inquiries about academic and sponsoring opportunities at poly(at)pharma.ai.


In collaboration with:
DNNs were trained on NVIDIA GPU


We thank NVIDIA for providing valuable GPU equipment for deep learning to Insilico Medicine.