OmicFormer: a statistical priors–informed transformer for accurate and generalizable omics prediction of diseases and complex traits
Preprint
- 4 August 2026
- preprint
- Published by Springer Nature in Research Square
Abstract
Precision medicine faces a critical challenge in translating high-dimensional omics data into robust disease predictions across diverse populations. Current approaches often fail under distribution shifts, partly due to their inability to encode complex biological feature dependencies. We present OmicFormer, a Transformer-based architecture that embeds two complementary statistical priors, i.e., feature–label associations and feature–feature dependencies, directly into its representation learning. This design captures local and long-range omic interactions often missed by conventional methods. Analyzing 500,000 UK Biobank participants, OmicFormer significantly outperforms strong baselines across 450 disease and 900 trait prediction tasks, with substantial gains spanning diverse metabolic, neurological, cardiovascular, and gastrointestinal conditions, alongside enhanced prediction of circulating metabolites, bone density traits, and retinal imaging biomarkers. Crucially, OmicFormer demonstrates robust generalization, achieving a substantial improvement over tree-based methods in an independent proteomics cohort across 19 diseases (GNPC, N = 7,289), and outperforming tree-based models across 50 multi-site neuroimaging sites (N = 4,728) for autism and schizophrenia classification. By explicitly embedding statistical structure, OmicFormer provides an interpretable and generalizable foundation for omics-based precision medicine.Keywords
All Related Versions
This publication has 22 references indexed in Scilit:
- Plasma proteomic signatures of social isolation and loneliness associated with morbidity and mortalityNature Human Behaviour, 2025
- Interpretable discovery of patterns in tabular data via spatially semantic topographic mapsNature Biomedical Engineering, 2024
- Plasma proteomic profiles predict individual future health riskNature Communications, 2023
- Tabular data: Deep learning is not all you needInformation Fusion, 2022
- Using deep learning to model the hierarchical structure and function of a cellNature Methods, 2018
- Identification of autism spectrum disorder using deep learning and the ABIDE datasetNeuroImage: Clinical, 2018
- Development and Validation of a Deep Learning System for Diabetic Retinopathy and Related Eye Diseases Using Retinal Images From Multiethnic Populations With DiabetesJAMA, 2017
- Natriuretic peptides and integrated risk assessment for cardiovascular disease: an individual-participant-data meta-analysisThe Lancet Diabetes & Endocrinology, 2016
- The autism brain imaging data exchange: towards a large-scale evaluation of the intrinsic brain architecture in autismMolecular Psychiatry, 2013
- Multitask LearningMachine Learning, 1997