Accuracy, efficiency and robustness of four algorithms allowing full sibship reconstruction from DNA marker data

Abstract

In the problem of reconstructing full sib pedigrees from DNA marker data, three existing algorithms and one new algorithm are compared in terms of accuracy, efficiency and robustness using real and simulated data sets. An algorithm based on the exclusion principle and another based on a maximization of the Simpson index were very accurate at reconstructing data sets comprising a few large families but had problems with data sets with limited family structure, while a Markov Chain Monte Carlo (MCMC) algorithm based on the maximization of a partition score had the opposite behaviour. An MCMC algorithm based on maximizing the full joint likelihood performed best in small data sets comprising several medium-sized families but did not work well under most other conditions. It appears that the likelihood surface may be rough and presents challenges for the MCMC algorithm to find the global maximum. This likelihood algorithm also exhibited problems in reconstructing large family groups, due possibly to limits in computational precision. The accuracy of each algorithm improved with an increasing amount of information in the data set, and was very high with eight loci with eight alleles each. All four algorithms were quite robust to deviation from an idealized uniform allelic distribution, to departures from idealized Mendelian inheritance in simulated data sets and to the presence of null alleles. In contrast, none of the algorithms were very robust to the probable presence of error/mutation in the data. Depending upon the type of mutation or errors and the algorithm used, between 70 and 98% of the affected individuals were classified improperly on average.

A graphical approach to relatedness inference

“…Summary of results (6) and (7) for experiments E1 and E2 applied to 8 test pedigrees using 8 loci. Summary of results (6) and (7) for experiments E1 and E2 applied to 8 test pedigrees using 10 loci. Summary of results (6) and (7) for experiments E1 and E2 applied to 8 test pedigrees using 12 loci.…”

“…This is manifest, for example, in the tendency of likelihood scores to split true sibling groups (Thomas and Hill 2000; Butler 2004). Alternative scores, such as the Simpson's Index proposed in Butler (2004) (see also Konovalov 2005), favor larger sibling groups and are therefore better able to preserve such family structure. This suggests that the selection of a method which properly accounts for the sample size and relationship complexity is crucial.…”

Heterozygote deficiencies in parasite populations: an evaluation of interrelated hypotheses in the raccoon tick, Ixodes texanus

“…PEDIGREE uses an MCMC algorithm to sample the space of possible partitions of individuals, and calculates for each partition a score based on the pairwise score method described by Smith et al (2001) and Butler et al (2004). In order to identify the best partition of individuals into putative kin groups, we performed five PEDIGREE runs for each CP (500 000 iterations per run) with the following parameters: 'full-sib constraint' ¼ 0; 'temperature' ¼ 10; and 'weight' ¼ 1 (the functions of each of these three parameters are detailed in Supplementary Table S1).…”

Parentage and sibship exclusions: higher statistical power with more family members

“…Recently, various statistical methods have also been proposed to infer sibships in a one-generation sample of individuals, using the genotypes of these individuals at a number of marker loci (Painter, 1997;Almudevar and Field, 1999;Hill, 2000, 2002;Smith et al, 2001;Beyer and May, 2003;Wang, 2004;Butler et al, 2004). These methods allow the inference of sibships consisting of an arbitrary number of individuals by exclusion (Smith et al, 2001;Butler et al, 2004) or likelihood (Thomas and Hill, 2000;Wang, 2004) approaches.…”