Using Pre-training and Interaction Modeling for ancestry-specific disease prediction in UK Biobank

Thomas Le Menestrel, Erin Craig, Robert Tibshirani, Trevor Hastie, Manuel Rivas
4/26/2024
cs.LGq-bio.QMstat.AP

No official code found. Best implementation: rivas-lab/AncestryOmicsUKB (1 star, last commit 2025-10-16).

Abstract

Recent genome-wide association studies (GWAS) have uncovered the genetic basis of complex traits, but show an under-representation of non-European descent individuals, underscoring a critical gap in genetic research. Here, we assess whether we can improve disease prediction across diverse ancestries using multiomic data. We evaluate the performance of Group-LASSO INTERaction-NET (glinternet) and pretrained lasso in disease prediction focusing on diverse ancestries in the UK Biobank. Models were trained on data from White British and other ancestries and validated across a cohort of over 96,000 individuals for 8 diseases. Out of 96 models trained, we report 16 with statistically significant incremental predictive performance in terms of ROC-AUC scores (p-value < 0.05), found for diabetes, arthritis, gall stones, cystitis, asthma and osteoarthritis. For the interaction and pretrained models that outperformed the baseline, the PRS score was the primary driver behind prediction. Our findings indicate that both interaction terms and pre-training can enhance prediction accuracy but for a limited set of diseases and moderate improvements in accuracy

View on arXivView PDF

Code Implementations(1)

rivas-lab/AncestryOmicsUKBCommunity100%

Using Pre-training and Interaction Modeling for ancestry-specific disease prediction in UK Biobank (PLOS One)

10RSep 1, 20251 years ago

Similar Papers

Multi-Disease Deep Learning Framework for GWAS: Beyond Feature Selection Constraints

79% match2025

Training Flexible Models of Genetic Variant Effects from Functional Annotations using Accelerated Linear Algebra

79% match1 repo102025

Comparative Analysis of Data Preprocessing Methods, Feature Selection Techniques and Machine Learning Models for Improved Classification and Regression Performance on Imbalanced Genetic Data

78% match1 repo2024

Investigating the Impact of Histopathological Foundation Models on Regressive Prediction of Homologous Recombination Deficiency

78% match2026

How Effectively Can Large Language Models Connect SNP Variants and ECG Phenotypes for Cardiovascular Risk Prediction?

78% match2025

DYNA: Disease-Specific Language Model for Variant Pathogenicity

78% match2024

Cite this paper

@article{menestrel2024using,
  title  = {Using Pre-training and Interaction Modeling for ancestry-specific disease prediction in UK Biobank},
  author = {Thomas Le Menestrel and Erin Craig and Robert Tibshirani and Trevor Hastie and Manuel Rivas},
  year   = {2024},
  eprint = {2404.17626},
  archivePrefix = {arXiv},
  url    = {http://arxiv.org/abs/2404.17626v2}
}

Add a badge to your README

Shows whether official code is linked for this paper.

Code of Paper badge
[![Code of Paper](https://codeofpaper.com/badge/2404.17626.svg)](https://codeofpaper.com/paper/2404.17626)

Discussion