ClonEstiMate, a Bayesian method for quantifying rates of clonality of populations genotyped at two‐time steps

Abstract

Partial clonality is commonly used in eukaryotes and has large consequences for their evolution and ecology. Assessing accurately the relative importance of clonal vs. sexual reproduction matters for studying and managing such species. Here, we proposed a Bayesian approach, ClonEstiMate, to infer rates of clonality c from populations sampled twice over a short time interval, ideally one generation time. The method relies on the likelihood of the transitions between genotype frequencies of ancestral and descendent populations, using an extended Wright-Fisher model explicitly integrating reproductive modes. Our model provides posterior probability distribution of inferred c, given the assumed rates of mutation, as well as inbreeding and selfing when occurring. Tested under various conditions, this model provided accurate inferences of c, especially when the amount of information was modest, that is low sample sizes, few loci, low polymorphism and strong linkage disequilibrium. Inferences remained robust when mutation models and rates were misinformed. However, the method was sensitive to moderate frequencies of null alleles and when the time interval between required samplings exceeding two generations. Misinformed rates on mating modes (inbreeding and selfing) also resulted in biased inferences. Our method was tested on eleven data sets covering five partially clonal species, for which the extent of clonality was formerly deciphered. It delivered highly consistent results with previous information on the biology of those species. ClonEstiMate represents a powerful tool for detecting and inferring clonality in finite populations, genotyped with SNPs or microsatellites. It is freely available at https://www6.rennes.inra.fr/igepp_eng/Productions/Software.

New insights into the population genetics of partially clonal organisms: When seagrass data meet theoretical expectations

“…First, the mean R gradually increases with decreasing longevity and increasing module turnover, that is from P. oceanica to Z. noltii (Figure 1), and D * follows a similar trend despite exhibiting a narrower range of values. Although unable to provide a reliable quantitative assessment of the rates of clonality due to large subsampling bias (Arnaud‐Haond, Duarte, et al, 2007; Becheler et al., 2017; Gorospe et al., 2015; Stoeckel et al, 2019), R may remain useful in comparative studies and shows a broader range of discriminable scenarios than D *. With a comparable sampling scheme and effort, R (and D *) can help assess the relative importance of clonality among natural populations.…”

“…indices based on the number of MLGs or MLLs) richness ( R ) and diversity. Only two recent methods allow rates of clonality to be quantified from genetic data: CloNcaSe (Ali et al., 2016), developed for organisms with cyclical parthenogenesis, and ClonEstiMate (Becheler et al., 2017), a Bayesian method relying on temporal samplings ideally separated by one generation. Nevertheless, their use relies on the important but seldom met requirement of temporal sampling and rather precise knowledge of generation time.…”

After a catastrophe, a little bit of sex is better than nothing: Genetic consequences of a major earthquake on asexual and sexual populations

“…For farms, where asexual reproduction has been reported (Guillemin et al, 2008), they were estimated for incremental values of c (the rate of clonal reproduction): 0, 0.1, 0.2, 0.3, 0.4, 0.5, 0.6, 0.7, 0.8, 0.9, 0.95, 0.98, 0.99, 0.999, and 1 (Becheler et al, 2017 Note: The status of each population is described in the column "effect EQ." For impacted populations, we distinguished three substatus, before earthquake (status before), directly after it (status EQ), and 2 years after (status after).…”

Expansion of frozen hybrids in the guppy ectoparasite,Gyrodactylus turnbulli

“…The disadvantage of recombinants may stem from restoring homozygotes for recessive deleterious mutations. Discrimination between these two possibilities ideally requires direct estimates of recombination rate, possibly based on changes in genotype frequencies over time in both hybrid and non-hybrid populations (Becheler et al, 2017).…”