ClonEstiMatePoly tab
This tab allows users to compute the posterior probabilities of joint
rates of clonality and selfing in polyploid populations genotyped at, at
least, two-time steps. This method was demonstrated to be the most
accurate way to quantitatively assess reproductive modes in diploid
populations over multiple Eukaryotes species, especially for detecting
low rates of clonality (Becheler et al. 2017). It should facilitate the
detection of clonal reproduction, the estimation of the rates of
clonality in polyploid populations, and promote the study of
reproductive modes and their genetic consequences in such species. It
should be a nice addition to the method of estimation of selfing rates
using multilocus standardized identity disequilibrium coefficient found
in spagedi (Hardy 2016).
Here, we extended to autopolyploids the Bayesian formula and method
ClonEstiMate from Becheler et al. (2017). It exploits the
likelihood of transitions of genotype frequencies from one generation to
another to accurately estimate rates of mutation, clonality and selfing,
and thus works well even in the absence of equilibrium between
evolutionary forces (genetic drift, mutation and rates of clonality)
which is quite common in partially clonal populations (Reichel et al.
2016). This method remains accurate using from about ten polymorphic
markers, even physically linked and mutating with other mutation model,
and from 30 sampled individuals. It is however sensitive to erroneous
assumed or restricted prior values of clonal and selfing rates, null
alleles and sampling time interval greater than two generations.
Extended equations for autopolyploids can be found in the documentation
in supplementary material. This discretized Bayesian method needs an
analysis plan listing discretized priors on rates of mutation, clonality
and selfing for each population (Fig. S1). Restricted ranges of prior on
each of these parameters allows better inferences on other targeted
parameters. Analysis plan can be uploaded or prepared (and saved for
future use) using the graphical interface. Analysis plan can be browsed
and checked using the integrated browser before launching the
computations. To speed-up the calculations, computations per locus and
population of the analysis plan were parallelized using the maximum
number of threads available by the operating system. Results are stored
in the folder containing GenAPoPop in a text-file separator
tabulation file that can be readily handled using any spreadsheet
application. Results are presented per population between two time-steps
as a list of discrete joined values of mutation rate, rates of clonality
and selfing with the corresponding posterior probabilities of such
joined combination of priors. This presentation of the results makes it
easy to combine the posterior probability mass functions per population
and generations into table and/or into plots of their distributions. If
found in the dataset, it also returns the list of monomorphic loci at,
at least, one sampling time. Monomorphic loci decrease the inference
power of the dataset to assess rates of mutation, clonality and selfing
between the two sampled generations.