3 ms·
> How else would you test the safety of a new drug or other chemical? Use computer models first as much as possible, while we know they are imperfect. In vitro
by ekianjo 5y ago
> How else would you test the safety of a new drug or other chemical?
Use computer models first as much as possible, while we know they are imperfect.
In vitro-studies before in-vivo. Then test on volunteers (paid), starting with very very small doses of whatever you want to test to reduce the risk to the maximum, and measure the effects on living systems as you go to better understand how it is metabolized.
- refurb 5y agoAs someone who has done drug development this is a horrific idea. I remember using computer modeling - the lead said “oh yeah, tweak the molecule here and you’ll double the potency”. Cool, make the tweak and…potency was 1/10th. And we do toxicity filtering already, before we get to animal models (why trying something you know will fail?). Despite this we have plenty of highly toxic molecule make it through. I remember the rats who testicles went necrotic after the first dose. Never saw that coming and thank god it didn’t happen in humans. We’re so far away from even modeling molecules docking with receptors. Toxicity? You’d have plenty of phase 1 volunteers dying every year. Or being horribly harmed.
- foobar33333 5y ago>Use computer models first as much as possible Are we not doing this already? I would assume this is even cheaper than producing the thing and testing it. And I'm not sure how confident I would be with computer models since we don't know a whole lot about how biology works enough to simulate it (given the simulation would mostly be trying to replicate previous potentially flawed studies)
- nceqs3 5y agoYou are correct, if you could build a computer model that could even somewhat predict the efficacy of a drug you would instantly become the wealthiest person in the world.
- foobar33333 5y agoI imagine the only real way to do such a thing would be to simulate physics on an atomic level and have a perfect atomic map of the body. Which we are no where near achieving yet. But once we have that sorted, all kinds of incredible simulations could be done.
- refurb 5y agoYou would need to do that, but not just for the target in question but the entire human body - there are so many feedback loops and mechanisms in the body to counter whatever your trying to do. Add on top toxicity and that’s why biology is so hard. Reminds me of the story of Lyrica. Logic was it targeted GABA receptors - made sense too since it looked very similar to GABA itself. Sweet! And it had the same effect as a GABA agonist would. Pretty god damn smart people thought that one up! Then a few years after approval someone figured out it actually worked on glutamate receptors. Complete luck that it looked like GABA, behaved like a GABA agonist. It just didn’t work that way at all though.
- fabian2k 5y agoWhat computer models? I can imagine that it is possible to predict some already known toxicological mechanisms for similar molecules. But we're certainly not even close to anywhere where we could predict unknown toxicological interactions. And in vitro tests are very limited in that area. Of course they're still done and useful, but they can't replace animal tests. And testing on paid volunteers is still ethically problematic, you're targeting pretty desperate people if you remove animal experiments which could detect toxic compounds.
- nceqs3 5y agoThere is no computer model that is even close to being accurate enough. Science actually understands very little about the human body on a very micro scale.
- yakshaving_jgt 5y agoIIUC, you are proposing computers -> petri dishes -> people Whereas the status quo is computers -> petri dishes -> animals -> people So you think safety testing should be less rigorous?
- foobar33333 5y agoAnd that poor people who take up these positions should be the first line of defense in testing.
- PeterisP 5y agoSafety of people matters a lot, but I keep thinking about the following observation - in drug trials we routinely see many drugs that work in vitro, work in mice but don't work in humans, and many drugs that seem safe for mice but have bad side-effects for humans. Shouldn't this cut both ways symmetrically? Does that imply that we have a bunch of drugs developed for as-yet-uncured diseases that work in vitro and would actually work for humans but just won't work on mice, so we never tried them on humans and will never know because it's unethical to try?
- yakshaving_jgt 5y agoDo you work in this field? I don't, but even as a layman I'm quite certain that the research process is more involved than you appear to be suggesting. As far as I'm aware, trials in animal models don't stop at mice — I've heard of researchers testing for safety and efficacy on mice, and then progressing to guinea pigs and subsequently macaques. Some animal anatomies behave more similarly to humans when confronted with the same disease than others. I know what I'm an expert in and it certainly isn't this. I trust actual scientists to know what they're doing.