AI-designed COVID-19 drug nominated for preclinical trial

An oral medication designed by scientists with the help of AI algorithms could one day treat patients with COVID-19 and other types of diseases caused by coronaviruses. Insilico Medicine, a biotech startup based in New York, announced on Tuesday it had nominated a drug candidate for preclinical trials – the stage before you …

[M. T. Ness]

A chemical substance is not a drug

It is the equivalent of the hardware of a computer. A drug has a dosage and a dosage schedule, which is the software.

The first thing to have a plan for is what to do when the substance doesn't seem to work in a living animal, or only works at doses that kill the animal.

[Filippo]

Re: A chemical substance is not a drug

I think the plan for those cases is that you document the failure, and then throw everything out. Discovery of drugs is pretty much a process of trying hundreds of molecules and hoping that one of them does more good than harm to patients with a specific set of conditions.

The usefulness of machine learning in this fields is in pointing you towards the ones that look more likely to work. That's a far cry from "AI can design drugs", but it's still a useful tool. At least in principle; it remains to be seen whether it's actually effective in suggesting molecules that, at the end of the process, actually work.

[Anonymous Coward]

Re: A chemical substance is not a drug

Exactly what I used to tell the bouncers on a Saturday night

[Pascal Monett]

ROTM anyone ?

So it has started. Machines are now designing drugs to be used on humans.

[breakfast]

Re: ROTM anyone ?

Machines have been making the drugs we designed or found for a very long time and that seems to have worked alright. Anyways, we've been designing things to put into machines for ages, so it only seems fair.

[Snowy]

The next.

Only 10 to 15 years (on average) from preclinical trials to the drug being ready. This read more like a company looking for some funding than announcing a treatment.

[Throatwarbler Mangrove]

Re: The next.

There's something going on right now, exactly what it is escapes me at the moment, but it seems to have caused governments around the world to fast track the testing and approval process for SARS-Cov-2, so if this one works, I suspect we'll see approval rather sooner than that time frame.

[Richard 12]

Re: The next.

This one is about to start the stage more accurately known as "Try in a petri dish to see whether it does anything interesting at all".

Something like 99% of compounds are discarded at this point. Sadly, most of these negative results are never published so other researchers spend a lot of time re-testing the exact same failures...

If this AI can improve that preclinical failure rate then it becomes interesting, but until then it's no different to tossing random compounds at the wall to see if any stick.

[Anonymous Coward]

@Throatwarbler Mangrove - Re: The next.

Nah, it's just big pharma positioning themselves to extract a steady revenue stream. Here in Canada they have more subscribers than Netflix.

[Manolo]

Re: The next.

"governments around the world to fast track the testing and approval process"

That is called a "conditional marketing authorisation" for EMA and a "emergency use authorization" by the FDA.

Not unheard of and not be confused with regular approval.

[Anonymous Coward]

Question is

Will this be a treatment in case someone gets sick or a (mandatory by government courtesy) lifestyle drug, 4 to 8 doses per year for everybody 5 years and plus ?

I'm asking for a friend of course.

[Conundrum1885]

Re. IND

Hey, just a heads up. Wonder if any of the data from the folding@home Collective was used for this?

I'd be intrigued to see if so, what percentage of the usage etc.

Having helped with this project through both lockdowns at great personal expense.

[Korev]

Re: Re. IND

It's hard to know as the company would want to keep their technology secret, but my slightly educated guess* is that they didn't. I imagine that they used structures of the protease which were probably obtained by public sources.

* I work in research for another pharma company

[Anonymous Coward]

Oh how “wonderful”

It’s a shame that the motive is profit

I’ll take my chances thanks, and avoid the medical industrial complex as if my life depended on it

[Throatwarbler Mangrove]

Re: Oh how “wonderful”

"medical industrial complex"

You are very intelligent.

[Francis Boyle]

Thing is

people like to get paid for their work. The alternative is problematic. And yes the profit motive has it's problems it turns out to be a very good way to get people to take risks.

[Anonymous Coward]

@Francis Boyle - Re: Thing is

You mean a very good way to get other people to take risks.

[Conundrum1885]

Folding@home

Think I saw this mentioned on some project or other.

Nice to see that its now being tested thanks to everyone's hard work and GPU cycles.