Explaining the unexplained admixture mapping signals via rare variants: the HCHS/SOL

Overview

This study investigates the genetic basis of previously unexplained admixture mapping signals in admixed populations, specifically focusing on Hispanic/Latino populations. Admixture mapping identifies genomic regions where local ancestry associates with phenotypic outcomes, detecting signals when causal variants differ in frequency or effect across ancestral populations. Prior research demonstrated that adjusting for nearby common variants identified through genome-wide association studies does not fully account for some admixture mapping signals, suggesting that additional genetic factors contribute to these associations. The research examines two hypotheses to explain residual admixture mapping signals: the contribution of rare genetic variants and the possibility that common variants located in broader genomic regions beyond the initially examined loci account for the unexplained signals. The investigation utilizes whole-genome sequencing data integrated with metabolomics profiles from the Hispanic Community Health Study/Study of Latinos cohort.

Methods and approach

The analysis employed whole-genome sequencing data coupled with metabolomic measurements from participants in the Hispanic Community Health Study/Study of Latinos. The research tested two distinct approaches to explain admixture mapping signals that remained after adjusting for previously identified common variants from genome-wide association studies. The first approach involved systematic inclusion of sets of rare variants to determine whether low-frequency genetic variation could account for the residual signals. The second approach expanded the genomic regions under consideration when identifying common variants, examining whether causal common variants located at greater distances from the admixture mapping peak could explain the signals. The study conducted comprehensive assessment across multiple metabolite phenotypes, implementing replication strategies to validate associations between rare variant sets and metabolite levels.

Results

The investigation detected multiple sets of rare variants demonstrating replicated associations with metabolite levels across the study population. However, these rare variants collectively explained only a small fraction of the admixture mapping signals that remained unexplained by initially identified common variants. In contrast, the second approach of expanding the genomic search region for common variants revealed that common variants located in larger genomic intervals surrounding the admixture mapping peaks accounted for the majority of the previously unexplained admixture mapping signals. These findings indicate that the primary source of residual admixture mapping signals derives from common variants positioned at greater genomic distances from the initial association peaks rather than from contributions of rare variants with large effects.

Implications

The findings clarify the genetic architecture underlying admixture mapping signals in admixed populations and demonstrate that common variants distributed across broader genomic regions constitute the predominant explanation for signals not captured by localized genome-wide association study hits. The limited contribution of rare variants to admixture mapping signals suggests that the differential ancestry patterns observed in these regions primarily reflect allele frequency differences of common variants across ancestral populations rather than ancestry-specific rare variants with substantial effects. This has methodological implications for admixture mapping studies, indicating that comprehensive evaluation of common variant contributions across extended genomic intervals should precede or accompany investigations of rare variant effects. The results also inform the design of follow-up studies aimed at identifying causal variants underlying admixture mapping associations and suggest that fine-mapping efforts in admixed populations should prioritize thorough interrogation of common variation across broad regions rather than focusing exclusively on rare variants near mapping peaks.