This is already done. Essentially making a medicine goes like this: 1) is it functional? Yes -> step 2 2) is it cytotoxic? (kills everything, for example by d

This is already done. Essentially making a medicine goes like this: 1) is it functional? Yes -> step 2 2) is it cytotoxic? (kills everything, for example by dissolving cell membranes) No -> goto step 3 3) is it neurotoxic? (prevents neurons from working) No -> goto step 4 4 to 199) is it ...toxic? 200) proceed to in vitro testing Someone made a paper ... what happens if you just reverse steps 2 through 199? Start with any old protein, and keep changing it to display as many different kinds of toxicity as possible. Answer: This just works. Works very well, since we know computational tests for a whole lot of toxicities, and at least a few of them are differentiable. So you can just assume essentially every country has already done this and tested it. Fortunately this is not such a huge issue. The challenge is not making poisons, the challenge is spreading them effectively (without ...). Some of these proteins would at least need contact between a large drop of the protein and your skin. Most would need to be injected to kill you. So ... And we know plenty of non-protein molecules, much easier and faster to make, available in industrial quantities without much checking, that are about as bad as the worst of these proteins. Also: there are radioactive substances sold on Amazon that'll kill you slowly and painfully just being near you. https://unicri.it/sites/default/files/2021-12/21_dual_use.pdf https://unicri.it/sites/default/files/2021-12/21_dual_use.pd...