> racially-targeted bioweapon Possible with relatively low specificity and sensitivity (high false positives, high false negatives, respectively). But society'

> racially-targeted bioweapon Possible with relatively low specificity and sensitivity (high false positives, high false negatives, respectively). But society's concept of "race" is nearly exclusively limited to physical appearance and has no basis in any part of the genotype which doesn't affect visual phenotype. Therefore to achieve high specificity and sensitivity you'd need a bioweapon which targets individuals based on complex combinations of genes which together govern physical appearance, often in relatively subtle ways. This is relatively impossible because overall physical appearance isn't a vector for infections. Even if you did this relatively impossible thing, you'd get a lot of false positives and negatives for most populations due to intermixing -- Latin America for example has very, very few groups of indigenous populations which aren't significantly mixed with the European colonizers' genes. When I say it's "possible with relatively low specificity and sensitivity" I mean that you could find other genes that are somewhat loosely correlated with our current definitions of "race", such as how some Africans are somewhat more likely to have sickle-cell disease (SCD) and that provides about a 75% protection against HIV, and heterogenous carriers of the gene (SCT - Sickle Cell Trait) have about 40% lower incidence rate of HIV. But even in the most concentrated areas of Africa, only 3% of some populations have SCD. And that dismal math will repeat for any "race-based" infection system -- you might be able to get slightly higher rates of infection in another target "race" but it won't be dramatically different enough from the effects on your constituent population to make it useful.