GenomEn
Risk and Association beyond Linearity
A First Step Towards Closing the Missing Heritability Gap
Phenotype predictions demo

Built to model gene-gene intercations (incl. X)

GenomEn is an ensemble of (non-linear) estimators that allows to model gene-gene interactions that traditional PRS methods neglect. It natively integrates the X sex chromosome.

Get started View API

Inspect GenomEn's predictions and explore learned associations

We share summary statistics, predictions, variant importance scores, and intra-trait architectures derived from GenomEn models for various phenotypes. Model artifacts are also available for download.

Browse Phenotypes →
Example of Interaction Manhattan Plot
Phenotype predictions demo

Make your own predictions with the genomen package

Use the genomen python package to train your own models, make phenotype predictions, and explore complex traits via GenomEn's variant importance values.

See docs →
Explore GenomEn

Documentation

Understand how you can make your own phenotype predictions with the ready-to-use genomen package.

Browse Phenotypes

Browse phenotypes and explore GenomEn's predictions, associations, and interactions

Read The Paper

Customizable and fast pricing page, integrated into the billing portal.

Download Artifacts

Download pre-trained model artifacts and variant effect sizes.

BibTeX

@inproceedings{Thomassin2025GenomEn,
      title = {Polygenic risk and association beyond linearity},
      author = {C. Thomassin, M. Franquesa Mones, D. Bonet, P. A. Gerlach, M. Comajoan Cara, D. Mas Montserrat, and A. G. Ioannidis},
      booktitle = {bioRxiv Preprint},
      year = {2025},
      note = {Preprint},
      url = {https://your-site.com/genomen}
    }