Abstract Body: Introduction:
Congenital heart disease (CHD) affects nearly 1% of live births and is the most common birth defect. However, interpretation of disease-associated noncoding variants remains a major challenge. More than 90% of the de novo variants identified by human genetic studies lie outside coding regions, making it difficult to define their regulatory functions, target genes, and developmental consequences.
Aim:
We developed AlphaOrganoid, an integrated framework combining AlphaGenome-based sequence prediction with human cardiac organoid assays to identify, prioritize, and functionally evaluate noncoding CHD-associated variants while nominating candidate target genes and novel risk genes.
Method:
We applied AlphaGenome to annotate CHD-associated noncoding variants and predict allele-specific effects on regulatory features relevant to cardiogenesis, including chromatin accessibility, transcription factor binding, and gene expression potential. Variants were linked to candidate target genes and prioritized based on predicted disruption of cardiac developmental programs. To assess function, we used human induced pluripotent stem cell–derived cardiac organoids that model key features of early heart development in a multicellular context. Selected variants were evaluated for effects on enhancer activity, downstream transcriptional output, and cardiac lineage-associated phenotypes.
Results:
AlphaGenome-based annotation prioritized a subset of CHD-associated noncoding variants predicted to perturb cardiac regulatory elements and alter genes involved in heart development. These annotations also enabled nomination of candidate target genes and identification of putative novel CHD risk genes not captured by nearest-gene assignment alone. Initial functional studies in human cardiac organoids supported measurable regulatory effects for a subset of prioritized variants, consistent with computational predictions. Together, these findings show that this pipeline enriches for noncoding alleles with plausible developmental impact and helps focus downstream mechanistic studies.
Conclusion:
AlphaOrganoid provides a scalable framework for linking noncoding genetic variation to developmental gene-regulatory mechanisms in CHD. By integrating computational prioritization with functional testing in human cardiac organoids, this platform supports identification of candidate pathogenic regulatory alleles, improves target gene assignment, and accelerates discovery of novel CHD risk genes.