Gene Embedding Hub: a standardized resource for human gene representations
- 1ETH Zurich
Abstract
Gene embeddings are vector representations of gene identity that provide machine-learning models with information from diverse biological sources. However, their use is hindered by a lack of standardization: inconsistent identifiers, formats, and preprocessing conventions. In this work, we present a unified resource of 67 human gene-embedding sets derived from protein, DNA and RNA sequences, biomedical text, bulk expression data, single-cell expression models, functional data, biological networks, knowledge graphs, regulatory genomics and multimodal models, covering up to 20,097 GENCODE-annotated protein-coding genes. The collection includes both published precomputed embeddings and representations generated or extracted from publicly available models and biological data. Each embedding set is organized in a folder containing the embedding matrix, the Ensembl gene identifiers, machine-readable metadata, and processing documentation. We validate the resource by measuring global and local similarity, cross-embedding consensus, and cross-embedding predictability. The collection is designed to allow community-contributed embeddings that meet predefined technical and documentation requirements, enables comparison of gene representations across modalities and provides standardized embeddings for training and downstream applications.
Related
- OMICON: a community resource for studying gene coexpression networks in normal and neoplastic human brain samples
- Do protein representations encode variant function? Diagnosing residue-level protein attributions via embedding-space gradients
- GSET: Gene Signals from Embedding Trajectories select compact contexts for mechanistic GRN inference
- Large-Scale Network Embedding Predicts Opposing Functions of Adjacent Regions Within Human Prefrontal Cortex
- Standardized human tests show that most vision models are face-blind
BibTeX
@misc{tabares2026gene,
title = {{Gene Embedding Hub: a standardized resource for human gene representations}},
author = {D. Tabares and M. Bohl and V. Boeva},
howpublished = {bioRxiv},
year = {2026},
month = oct,
doi = {10.64898/2026.10.29.762818},
url = {https://www.biorxiv.org/content/10.64898/2026.10.29.762818v1},
note = {CC BY},
}
RIS
TY - UNPB
TI - Gene Embedding Hub: a standardized resource for human gene representations
AU - D. Tabares
AU - M. Bohl
AU - V. Boeva
PY - 2026
DA - 2026/10/29
PB - bioRxiv
DO - 10.64898/2026.10.29.762818
UR - https://www.biorxiv.org/content/10.64898/2026.10.29.762818v1
AB - Gene embeddings are vector representations of gene identity that provide machine-learning models with information from diverse biological sources. However, their use is hindered by a lack of standardization: inconsistent identifiers, formats, and preprocessing conventions. In this work, we present a unified resource of 67 human gene-embedding sets derived from protein, DNA and RNA sequences, biomedical text, bulk expression data, single-cell expression models, functional data, biological networks, knowledge graphs, regulatory genomics and multimodal models, covering up to 20,097 GENCODE-annotated protein-coding genes. The collection includes both published precomputed embeddings and representations generated or extracted from publicly available models and biological data. Each embedding set is organized in a folder containing the embedding matrix, the Ensembl gene identifiers, machine-readable metadata, and processing documentation. We validate the resource by measuring global and local similarity, cross-embedding consensus, and cross-embedding predictability. The collection is designed to allow community-contributed embeddings that meet predefined technical and documentation requirements, enables comparison of gene representations across modalities and provides standardized embeddings for training and downstream applications.
ER -