Análisis de parentescos mediante el uso de marcadores moleculares

Resumo

The inference of parentage relationships between individuals startingfrom their molecular resemblance represents an useful methodology forgenetic improvement in Aquaculture. This makes possible the reorganization of broodstocks in groups with low relatedness, as well as the identification offamilies in selection programs to avoid the harmful effects of inbreeding andto maintain the highest diversity as possible across generations. The choiceof the appropriate genetic marker and the statistical methodology is essen-tial to get the best solution for the questions considered. Depending on theavailability of previous information on the genealogy, paternity or parentageanalysis will be applied. The simplest approach to paternity inference, invol-ves parent identification through the exclusion of the remaining candidates.However, this can lead to more than one solution, solvable after applicationof maximum likelihood procedures. The ability to identify the parents of eachoffspring depends on the sampling scenario (number of candidate parentsand the fraction sampled), on the potential of the markers used (polymor-phism, genotyping errors), and on the conformance to theoretical assumptio-ns of the statistical models applied. A large number of paternity programsare freely available. These usually present complementary performances, andthere is not the best software for all situations.When we deal with a single group of individuals belonging to the samegeneration or that can not be separated into known generations, the aimis just to estimate the degree of genetic relationship between them, usuallyexpressed as the coancestry coefficient. The basic idea is to determine howmuch of the molecular similarity (Identity By State) is due to the IdentityBy Descent (the really important parameter). There are two groups of rela-tionship estimators from the molecular information. One of them includesmethods directed at estimating coancestry for only a pair of individuals ata time, usually relaying on the knowledge of the allele frequencies of thereference population. Estimators within this group can be further dividedinto those called Method of Moments Estimators (MME) and those basedon Maximum Likelihood (MLE). The other group of methods uses jointly theinformation of all individuals to determine the more probable population/familiar structure. They perform an explicit reconstruction of the genealogyleading to the observed population (at least for one generation above). Depen-ding on the type of estimator and the assumptions used in their development,each of them presents some advantages and/or limitations which are to betaken into account when choosing the one to use. In the present study, themost relevant estimators are presented and, also, a list of free software whichimplement them is made available