Modified SLCA algorithm for full-sib reconstruction of haplodiploid species

Abstract

Full-sib reconstruction from molecular marker data in the absence of parental information is an active research area. Such analyses can provide useful information for studies of mating systems and gene flow, and for estimating effective population size in the wild. Although various methods have been proposed, but their estimation accuracies for some applications are not known. Here we propose a modified version of Shared Loci Correspondence Analysis (mSLCA) to reconstruct full-sib families of haplodiploid species. We ran simulations to compare the accuracies of the original SLCA and mSLCA. mSLCA outperformed SLCA at various data settings. mSLCA produced accurate estimates of the number of full-sib families when the numbers of loci and alleles per locus were equal to or more than eight. We also used actual DNA data of commercial Bombus terrestris colonies to verify the validity of estimates with an increasing number of colonies. mSLCA outperformed SLCA at various colony sizes with slight underestimation. © 2007 Entomological Society of America.