AI Invents New 'Recipes' For Potential COVID-19 Drugs (sciencemag.org) 28
sciencehabit writes: As scientists uncover drugs that can treat coronavirus infections, demand will almost certainly outstrip supplies -- as is already happening with the antiviral remdesivir. To prevent shortages, researchers have come up with a new way to design synthetic routes to drugs now being tested in some COVID-19 clinical trials, using artificial intelligence (AI) software. The AI-planned new recipes -- for 11 medicines so far -- could help manufacturers produce medications whose syntheses are tightly held trade secrets. And because the new methods use cheap, readily available starting materials, licensed drug suppliers could quickly ramp up production of any promising therapies. "If you are going to supply a drug to the world, your starting materials have to be cheap and as available as sugar," says Danielle Schultz, a chemist at Merck. The new method, posted as a preprint this week, "is really solid," she says. "I am impressed by the speed at which [the researchers] were able to find new solutions for making existing drugs."
Patents give pharmaceutical companies the right to be the sole supplier of a new drug in a given country, usually for 20 years. Once a drug goes off patent, other companies can produce and sell it as a generic. The method to make the drug is often secret to discourage competition even after patents expire. But COVID-19 has changed all that, Schultz says. "We are at a time when it's all hands on deck." Only two medicines -- remdesivir and dexamethasone -- are currently proven to fight COVID-19. That has led to supply shortages for both. On 4 August, attorneys general from 34 U.S. states wrote federal officials, calling remdesivir supplies "dangerously limited," and urging states be given "march-in rights" to violate owner Gilead Sciences' patents. Such rights would allow states to work with third-party manufacturers to make additional supplies of the drug. To prevent future supply crunches, University of Michigan chemist Timothy Cernak and colleagues turned to a commercial drug synthesis AI program called Synthia. The software can help pharmaceutical manufacturers find the most efficient and cost-effective strategy for synthesizing medicines, most of which are fairly complex molecules that can be built in myriad ways -- much as an artist can apply brush strokes in infinite combinations to paint the same landscape. "It's more options than the human mind can comprehend," Cernak says.
Patents give pharmaceutical companies the right to be the sole supplier of a new drug in a given country, usually for 20 years. Once a drug goes off patent, other companies can produce and sell it as a generic. The method to make the drug is often secret to discourage competition even after patents expire. But COVID-19 has changed all that, Schultz says. "We are at a time when it's all hands on deck." Only two medicines -- remdesivir and dexamethasone -- are currently proven to fight COVID-19. That has led to supply shortages for both. On 4 August, attorneys general from 34 U.S. states wrote federal officials, calling remdesivir supplies "dangerously limited," and urging states be given "march-in rights" to violate owner Gilead Sciences' patents. Such rights would allow states to work with third-party manufacturers to make additional supplies of the drug. To prevent future supply crunches, University of Michigan chemist Timothy Cernak and colleagues turned to a commercial drug synthesis AI program called Synthia. The software can help pharmaceutical manufacturers find the most efficient and cost-effective strategy for synthesizing medicines, most of which are fairly complex molecules that can be built in myriad ways -- much as an artist can apply brush strokes in infinite combinations to paint the same landscape. "It's more options than the human mind can comprehend," Cernak says.
Bad news (Score:5, Funny)
Re: (Score:2, Insightful)
Re:Bad news (Score:4, Insightful)
and there's evidence for it helping treat other coronaviruses.
No, there isn't. Stop lying. You're not up to the level of the con artist.
We tested it, and it turned out not to be useful.
And that is where the discussion ends. Tests were done and they found no benefit, and some significant downsides, to using it.
Re: (Score:2)
You're being foolish. It was a reasonable thing to try. It was tried. It didn't work. If you didn't test it, you wouldn't have a reason to say it wouldn't work.
OTOH, it wasn't a *very* likely drug, just not an unreasonable one.
Re:Time Travel Trump (Score:2)
> and there's evidence for it helping treat other coronaviruses.
> No, there isn't. Stop lying. You're not up to the level of the con artist.
I'm with quonset, clearly Trump has acquired time travel technology and used it to plant misinformation in 2005. He's a time fascist!!!
https://pubmed.ncbi.nlm.nih.go... [nih.gov]
Wait, now 2004. I swear this didn't exist 5 minutes ago. He's actively changing history. This is worse than book burning.
https://pubmed.ncbi.nlm.nih.go... [nih.gov]
Re: (Score:2)
It's standard software. (Score:4, Interesting)
There is a lot of hype here because what they are describing is actually just run-of-the-mill software. You put in your desired molecule and it finds the most efficient paths to synthesizing it. Not only has it been used to get around patents but it's also been used to prevent there from being a way to get around patents. It's a double-edged sword of sorts.
Re: (Score:2)
Tightly held my ass (Score:4, Informative)
could help manufacturers produce medications whose syntheses are tightly held trade secrets.
Tightly held trade secrets my ass. The synthesis methods of medicinal drugs have to be submitted to national regulatory bodies and are *publically available*. Stop spreading FUD.
Re: (Score:3)
Maybe. Patents are supposed to reveal how to make whatever is patented, too. They're supposed to be sufficient that one "skilled in the art" could reproduce them. This is almost never the case. Are drug synthesis regulated more transparently?
For that matter, it's often the case that even when companies are trying to transplant a synthesis from one facility to another the documentation is incomplete enough that someone from the first factory has to come to the second factory to show them how to do it. A
Re: (Score:3)
And it can (has) happened that merely ordering a needed ingredient from a different supplier has been sufficient to cause the process to fail. (There was an unknown trace ingredient,
OT: I remember reading a story a decade or two ago (from The Register, I believe) about some semiconductor manufacturing line. They had the clean-room production line and setup, all of that song and dance. After a while they notice that one particular production line produced output markedly better than any of the other lines while using the exact same equipment and supplies. Wondering what was going on, they didn't tell the managers (to accidentally disturb what was happening) and surreptitiously placed
COVID-19 Drugs (Score:2)
Forget Disney with their 100-year and increasing Imaginary Mouse patents -- this might actually be something that you can't live without -- literally. (An offer you can't refuse?) Here we come, AEon Flux [wikipedia.org]
I welcome our new [SPOILER!] recycled-soul overlords.
AI Training. (Score:1)
Right now, AIs with random amounts of training should be considered way less valuable than even an uninformed person.
That is, about as good as an AI in a videogame.
It's easy to make an AI perform 'perfectly' under a set of canned circumstances - you can make an AI that can exactly target a person and kill them in a random game if that's your focus - that's the kind of mathematical results they're good at.
But they don't 'understand' the circumstances beyond what they're fed and can jumble together - and eac
Re: (Score:2)
Watson was overhyped, but it's probably still being developed and mark 2 will be more capable. (Still not up to the hype, of course.)
That said, multiple AIs are improving in their capability every year. Each year some of them become good enough to take over jobs that they weren't good enough to handle the prior year. The question is always "Are the ones that are good enough, cheap enough to run?". Sometimes the answer is "yes", and often they don't need to be anywhere near perfect to be "good enough".
Re: (Score:2)
It's got a "proven success rate" alright...and the rate is "worse than not giving any treatment". This doesn't mean everyone who's treated with it dies or is permanently incapacitated...but that's also true of those who aren't given any treatment.
Re: (Score:2)
"...so what do I know..."
Not nearly enough apparently. HCQ is not used "world wide" because its "proven success" is zero.
The taxpayers will still be shafted (Score:2, Insightful)
Even if AI found some recipe for a vaccine which was made from the three most common elements on the planet, taxpayers would still be told to bend over and suck it up after being forced to hand over billions of their dollars to a private company. Guaranteed the vaccine would still cost $50 per shot.
On a side note, whatever happened to Republicans saying if taxpayer money was used the organization receiving the money has to do what the government says? If we're forced to hand over billions of dollars of our
Re: (Score:2)
Are you backing that guarantee? If the cost per shot is less than $50, will you pay the difference between $50 and the cost, times the number of doses delivered, to some charity?
No? Then you're not making a guarantee.
Re: (Score:3)
We, the taxpayers, already paid for vaccine with the billions we were forced to hand over. Why should we have to pay a second time? What do you think this is, the insurance industry?
Re: (Score:2)
So here we are, months (at least) from knowing whether any of the vaccines currently in trial will be effective. Yet somehow you know not only that the effective vaccine will have been developed with taxpayer money, but you have also somehow managed to do a full cost accounting of development, testing, manufacturing, and distribution of that mystery vaccine and determined that taxpayer money has covered 100% of those costs. You must be a genius. Or an idiot.