You could try, but probably wouldn't want to be around when you released it. The prospect that rogue scientists could develop bioweapons designed to target certain ethnic groups based on their genetic differences was raised this week in a report by the British Medical Association (BMA).

The report, Biotechnology, Weapons and Humanity II, warns that construction of genetic weapons "is now approaching reality". Such "genetic bombs" could contain anthrax or bubonic plague tailored to activate only when genes indicated the infected person was from a particular group.

The report says the combination of human genome studies, the development of vectors capable of introducing harmful material to cells and new ways to disrupt genes should raise concerns about potential misuse.

Malcolm Dando, professor of peace studies at Bradford University and author of the BMA report, says: "The problem is that the same technology being developed to create new vaccines and find cures for Alzheimer's and other debilitating diseases could also be used for malign purposes."

The BMA, which dismissed the idea of genetic weapons in a 1999 report (Biotechnology, Weapons and Humanity I), has lifted its new concerns from the work of a German group called the Sunshine Project. It looked at how mutations in our genome called single nucleotide polymorphisms (SNPs) differ between specific ethnic groups and concluded: "Genome data in public databases revealed that hundreds, possibly thousands, of target sequences for ethnic specific weapons do exist. It appears that ethnic specific biological weapons may indeed become possible in the near future."

Rather than specifically triggering the toxic effects of organisms such as anthrax, the Sunshine project warned that weapons based on a new medical technique called RNA interference could shut down vital genes. If the sequence of the target gene varies between two different populations the technique could be used to interrupt key body functions in one population and not the other. "If as little as 10% or 20% of a target population would be affected, this would wreak havoc among enemy soldiers on a battlefield or in an enemy society as a whole," the group said.

Others say the concerns are exaggerated. "Trying to find a weapon that affects quite a few of one ethnic group and none of another ethnic group is just not going to happen," says David Goldstein, who studies population genetics at University College London. "Because all groups are quite similar you will never get something that is highly selective. The best you would probably do is something that kills 20% of one group and 28% of another."

The groups in question are also far broader than those associated with ethnic conflict. Geneticists can only distinguish between people with ancestry traced to regions such as Europe, Sub-Saharan Africa and East Asia.