Inbreeding and runs of homozygosity: A possible solution to an old problem

Abstract

Quantifying human, plant and livestock inbreeding has been an important goal of evolutionary biologists and agricultural scientists for nearly a century, and methods to do so continue to evolve. This review examines current approaches for estimating inbreeding at individual and population levels based on genetic information. Central to this approach is the detection of runs of homozygosity (ROH), long stretches of homozygous genome that most likely arise when the individual is the offspring of related individuals. When related individuals mate, the offspring carry long sections of the genome that are homozygous and identical by descent (IBD). Long ROH are most likely derived from a recent ancestor; shorter ones, from a more distant ancestor. Calculating how much an individual׳s genome occurs as ROH of particular lengths (e.g. >1Mb, >2Mb, and >4Mb) provides information about levels of inbreeding relative to reference populations specific numbers of generations ago. Although identifying and quantifying ROH can be complicated by genotyping errors and undetected heterozygosity within apparently continuous ROH, inbreeding estimates based on ROH clearly indicate that inbreeding levels in bovine and porcine populations are much higher than those in human populations. Frequencies of ROH vary widely within and across chromosomes, with chromosomes exhibiting ROH hotspots or “islands” as well as coldspots or “deserts”. The reasons for this variation are unclear and are attracting growing interest. Next-generation sequencing may improve our understanding of ROH and their usefulness as a tool in inbreeding research. We argue for combining ROH analysis and other genomic estimators unrelated to haplotype length in order to better define the inbreeding reference population.