Dear all,
We work with large families with mendelian inherited diseases. Parents of sick children are often related with each other. However, sometimes we do not know about the degree of relationship (e.g. second cousin or third cousin). Therefore we want to calculate inbreeding coefficients using SNP genotype data (AA, AB and BB). We use GenomeStudio. We tried Plink with `plink input report 2.1.3´ plug-in of GS but inbreeding coefficients had negative value. What do you think is the problem? Could you suggest another software or another way to calculate? Thanks.
Dear Nehir,
I have a long-term experience working in highly inbred genetic isolates for mapping genes of complex diseases. I using several methods for F evaluation in the isolates , in the pedigrees and based on genome scanned STRs and SNPs. Currently for SNPsw analyses I'm using Golden Helix SVS software that has much options, including HWE, coefficient of inbreeding etc. You can use it for evaluation of F (coefficient of inbreeding) based on SNPs. If you will get negative F -this says about larger number of observed heterozygotic subjects than expected. I have got such negative F as well but only in demographically younger genetic isolates (sec ondary isolates according Neel' (1992) definition)- negative F in this case says about 'younger' inbreeding (in latest generations) in comparison with positive F I observed in demographically ancient genetic isolates of indigeneous small ethnics that have traditional consanguneus marriages in their ancient history. Just one more note: F calculated by different methods will have different values, don't be confused with that. Hope, this information will be helpful for you. Best wishes
Dear Maureen,
Actually, we need the degree of relationship to create a pedigree file for linkage analysis. But using HWE for a quick check is definitely a great idea. We were surprised having negative values because we had expected some degree of relation. Thank you for the answer, we will try it.
Dear Kazima,
Thank you for the suggestion and explanations. We will try that software.
Hi Nehir,
Pedigree analysis is good and much accurate if done properly. And now runs of homozygosity much helpful. I have discussed with a researcher from Switzerland, who measured the expected risks in children by having parental DNA, surprisingly he found the higher estimated risk than in actual observations. So phenotype is also very much important and can't be ignored. Gene-environment interactions, epigenetic, and selection also operates simultaneously with inbreeding.
Thanks,
Fareed
Using pedigrees is not possible to calculate F-coefficient of inbreeding. Usually SNPs we are using hundreds thousands and more for whole genome scan. Only advanced software like of SVS from GoldenHelix etc have options for automatic evaluations of F. Interesting question can raise after evaluation of F -in some populations we can get negative values of F, in another ones -positive. But it is another question related with certain populations gene pool and demographic history that we can discuss later.
This is a somewhat old question, but I would like to ask if anyone can give a number for the very minimum number of SNPs needed to estimate a minimally reliable F. Thank you!