Thanks to everyone who wrote me suggesting possible explenations for the heterozytore excess that I encounter within some of my SNP loci. Some of these answers: Original massage Other: Heterozygote excess with SNP markers Dear EvolDir members, I'm working on the population genetics of seagrass, Zostera marina, using SNP loci. When I analyse the genotypic data for some natural populations I encounter high HWE departures. Some of the loci, especially some loci that are linked to each other in three different regions, have a high heterozygote excess. Some populations consist only of heterozygotes (sample size between 15-32). Based on the available msat data of these individuals, clones were excluded from SNP genotyping and subsequent data analysis. I'm not aware of any paper that encounters similar HWE departures. Does any of you encounter similar problems or does any of you have an idea what the cause could be? Any suggestions and tips will be greatly appreciated. Thanks Steven Ferber 1) Natalia M. Belfiore wrote: are you sure you're not genotyping multiple duplicated loci? When I see a locus with only hets (or nearly only), I assume the locus is duplicated and I'm amplifying 2 nearly identical copies of it. You could/should redesign the probe/genotype marker by extending it or moving it somehow to verify that you are only genotyping one copy of a locus. Natalia M. Belfiore nmb@berkeley.edu 510 643 0986 510 666 0314 2) Jeff Dole wrote: Steve, much of the literature going back decades show heterozygote excess/negative Fis, and the only reasonable interpretation that I know of is selection, either on the marker loci or others in linkage disequillibria with them. Identity disequillibria and selection against homozygotes in such disequilibria would not produce this effect, as even if only outcrosses survive, Fis would be zero, not negative. Jeff Jeff Dole Jepson Herbarium University of California Berkeley, CA 94720 3) Jeff Dole wrote: Steve, to follow up; I'm not familiar with the genetics of seagrass and the interpretation of SNP's, but with gene duplication (e.g., polyploidy) it's possible to have an excess of apparent heterozygosity (ranging to fixed heterozygosity) by having two or more duplicate loci fixed for different alleles. Jeff Dole Jepson Herbarium University of California Berkeley, CA 94720 4) Ellen M. Wijsman wrote: Most likely you are picking up more than one locus with the same assay. Where you have only heterozygotes, you most probably have 2 monomorphic loci, each with one of the alleles. Where you have an excess of heterozygotes, you might have contamination, or more than one locus, one of which might either be monomorphic (in which case one homozygote will be missing), or both of which are polymorphic for the same alleles. These issues are well understood in SNP discovery and use in human genetics. 5) Carolien de Kovel wrote: Beste Steven, Hoewel afwijkingen van HWE in mijn ervaring meestal technische fouten zijn, ga ik ervan uit dat je dat al hebt gecontroleerd. Hele stukken alleen heterozygoot klinkt vreemd, maar het deed me denken aan zelf-incompatibiliteit ( a paart alleen met alpha etc.). Is dat een mogelijkheid? Succes ermee, Carolien de Kovel 6) Anna Johansson wrote: Hi! One explanation could be that you have duplicated genomic regions so that you actually are genotyping two loci with two different fixed alleles. Anna Johansson Phd student, Lund university 7) Magdalena Zarowiecki wrote: Well, usually they say things like that can be caused by migration; if two different and divergent populations come together, that would cause heterozygote excess (the Wahlund effect). But I agree you case is extreme. Magdalena Zarowiecki 8) Monika wrote: Hi Steven, I do not have experience with SNPs and I do not know much of the life cycle of seagrass, so I cannot give you much of advice on that. However, I used to work on trematode parasites which had comlex life-cycle consisting of asexual and sexual stages, where we saw excess of heterozygotes (in terms of microsatelittes) regularly, even we excluded all clonal individuals. In any case, I would be interested in responses you get. Can you please summarize them and forward to me? cheers, Monika 9) Benoit Pujol wrote: Dear Steven, isn't Zostera having some clonal reproduction which could explain your high HW departure? Regards, Ben 10) Benoit Pujol wrote: Hi Steven, OK, but imagine one pool of clones that evolves towards fixing heterozygosity, which sounds logic. Now imagine a few differentiated clonal lineages which all fix heterozygosity but are all differentiated one from each other so that you obtain (A/B,C/D) and (E/F,G/H). Of course, I guess that a few generations of crosses among them will put all that at Hardy Weinberg unless mating isn't allowed between specific lineages. Another, maybe more 'naturally realistic' explanation would be that you have strong geitonogamy (selfing realized by crosses between clones) in your population and a very strong counter selection of inbred individuals early on. That's what happens in the plant species (cassava) I was doing my PhD on (Pujol et al 2005 ecology letters and Pujol and McKey 2006 JEB) but I don't know to what extent that could explain your heterozygosity excess. You can compute identity disequilibrium to visualize the observed heterozygosity distribution and the expected one and a simulated middle one. I believe that HWE is quite sensitive to geitonogamy+heterozygosity selection. A very slight directional destabilisation of the distribution has a major impact on the HWE departure statistics. Hope it helps, I look forward to see the summary of your findings about that question on the evoldir, All the best, Ben 11) Eric Normandeau wrote: Hi Steven, Duplicated parts of the genome (like duplicated chromosomes) could lead to such a very high proportion of "heterozygotes". In fact, you would be genotyping 2 different regions at the same time with variation at 1 nucleotide that would systematicly score as heterozygotes. I don't know your study organism so I cannot tell how likely this is the case, but it could very well be. Hope this can help you Cheers! Eric Normandeau S.Ferber@rug.nl