Exome Sequence Analysis and Coronary Artery Disease

Introduction

• Study Title: Exome sequence analysis identifies rare coding variants associated with a machine learning-based marker for coronary artery disease (CAD)
• Authors: Ben Omega Petrazzini et al.
• Published in: Nature Genetics, July 2024
• Institutions Involved: Icahn School of Medicine at Mount Sinai, University of Michigan, Mount Sinai Fuster Heart Hospital, among others
• DOI: 10.1038/s41588-024-01791-x

Abstract

• CAD is a complex disease with various risk factors and pathogenic processes.
• A machine learning-based in silico score for CAD helps capture progression, severity, and underdiagnosis.
• This study examines the association of rare coding variants with the in silico score.
• The analysis was conducted using data from UK Biobank, All of Us Research Program, and BioMe Biobank.
• Identified associations in 17 genes, with 14 showing moderate support for CAD.
• Observed a surplus of ultrarare coding variants in 321 aggregated CAD genes.
• The findings expand the genetic understanding of CAD and highlight the role of digital markers in genetic research.

Conflict of Interest

• R.D is a co-founder and consultant for various health companies.
• R.S.R has received research funding and consulting fees from multiple pharmaceutical companies.
• Declared unrelated patent applications and stock holdings by R.S.R.
• Remaining authors report no competing interests.

Figures

• Extended Data Fig. 1: Receiver operating characteristic (ROC) curves for ISCAD trained on different biobank datasets.
• Extended Data Fig. 2: Distribution of ISCAD scores in CAD cases and controls.
• Extended Data Fig. 3: Manhattan plot of rare coding variant association meta-analysis.
• Fig. 1: Study design schematic illustrating workflows for rare variant association studies.
• Fig. 2: Evidence for 17 genes associated with ISCAD in CAD biology.

Study Details

• Methods:
  • Used exome sequencing data and in silico scores from electronic health records.
  • Machine learning models fitted on data from UK Biobank, All of Us, and BioMe Biobank.
  • Rare variant association models explored using CAD status.

Gene Associations

• 17 Genes Identified:
  • Tier-1: Strong evidence from clinical trials, known associations, or drug effects.
  • Tier-2: Moderate evidence from CAD gene maps or eQTL signals.
  • Tier-3: Moderate evidence from associations with CAD risk factors.
  • Tier-4: Additional evidence from genome-wide significant associations.

Funding

• Supported by grants from the National Institutes of Health and the Department of Health & Human Services, USA.

Additional Resources

• Full text available from Nature Publishing Group and PubMed Central.
• MedlinePlus Health Information and NCI CPTAC Assay Portal for further information.

Related Research

• Enhancing prediction accuracy of CAD through machine learning.
• Studies on rare genetic variants and their impacts on various conditions.