Match on Family Finder full X

[John McCoy]

I hope people are not indulging in do-it-yourself imputation of DNA data! Imputed data isn't real data, it's a statistical guess about what the untested SNP's might be. The more SNP's are imputed, the greater the odds that some will be wrong, possibly enough to generate false matches or to mask real matches. Best to stick to the actual evidence rather than to introduce greater uncertainty.

[georgian1950]

I hope people are not indulging in do-it-yourself imputation of DNA data! Imputed data isn't real data, it's a statistical guess about what the untested SNP's might be. The more SNP's are imputed, the greater the odds that some will be wrong, possibly enough to generate false matches or to mask real matches. Best to stick to the actual evidence rather than to introduce greater uncertainty.

You are absolutely right, but I am worried about imputation on a professional level, too. When one goes from thousands of matches on Family Finder to a handful on DNA Land, imputation does not pass the "smell test". I am finding some bad assumptions in genetic genealogy autosomal methodology. Rather than forcing a solution with massive computational power, genetic genealogy needs to get the basics right. When you imputate with bad assumptions, the computer programs become massive GIGO processors. You will have a lot of wasted resources and efforts, not to mention confused and defrauded consumers of DNA testing.

Jack