Correspondence Between Genomic- and Genealogical/Coalescent-Based Inference of Homozygosity by Descent in Large French-Canadian Genealogies.

Kelly M Burkett,Ellen E Freeman,Marie-Hélène Roy-Gagnon,Mohan Rakesh,Catherine Laprise,Hélène Vézina,Patricia Morris

doi:10.3389/fgene.2021.808829

Abstract

Research on the genetics of complex traits overwhelmingly focuses on the additive effects of genes. Yet, animal studies have shown that non-additive effects, in particular homozygosity effects, can shape complex traits. Recent investigations in human studies found some significant homozygosity effects. However, most human populations display restricted ranges of homozygosity by descent (HBD), making the identification of homozygosity effects challenging. Founder populations give rise to higher HBD levels. When deep genealogical data are available in a founder population, it is possible to gain information on the time to the most recent common ancestor (MRCA) from whom a chromosomal segment has been transmitted to both parents of an individual and in turn to that individual. This information on the time to MRCA can be combined with the time to MRCA inferred from coalescent models of gene genealogies. HBD can also be estimated from genomic data. The extent to which the genomic HBD measures correspond to the genealogical/coalescent measures has not been documented in founder populations with extensive genealogical data. In this study, we used simulations to relate genomic and genealogical/coalescent HBD measures. We based our simulations on genealogical data from two ongoing studies from the French-Canadian founder population displaying different levels of inbreeding. We simulated single-nucleotide polymorphisms (SNPs) in a 1-Mb genomic segment from a coalescent model in conjunction with the observed genealogical data. We compared genealogical/coalescent HBD to two genomic methods of HBD estimation based on hidden Markov models (HMMs). We found that genomic estimates of HBD correlated well with genealogical/coalescent HBD measures in both study genealogies. We described generation time to coalescence in terms of genomic HBD estimates and found a large variability in generation time captured by genomic HBD when considering each SNP. However, SNPs in longer segments were more likely to capture recent time to coalescence, as expected. Our study suggests that estimating the coalescent gene genealogy from the genomic data to use in conjunction with observed genealogical data could provide valuable information on HBD.

Highlights

Inbreeding leads to increased homozygosity and has a negative effect on phenotypes (Charlesworth and Willis, 2009)
We simulated chromosomal segments using a coalescent model combined with genealogical data from two study samples from the French-Canadian founder population
We used these simulations to compare two genomic measures of homozygosity by descent (HBD) (FEstim and IBDLD) with HBD and relatedness measures based on the study and coalescent/gene genealogies

Summary

Introduction

Inbreeding leads to increased homozygosity and has a negative effect on phenotypes (Charlesworth and Willis, 2009). This phenomenon is referred to as inbreeding depression and is well documented in plants and animals (Roff, 1997). Most human populations display restricted ranges of homozygosity by descent (HBD), making the identification and quantification of the effect of inbreeding depression challenging in humans (Keller et al, 2011; Yengo et al, 2021). The FrenchCanadian founder population originated at the beginning of the 17th century with the immigration of French settlers (Charbonneau et al, 1993), which ended in 1759 after the British conquest. Urban regions like the Montreal region saw more diverse immigration patterns (McInnis, 2000; Piché, 2003), including migration from other regions of Quebec and more mixing

Objectives

Methods

Results

Conclusion