AI Plus a Chemistry Robot Finds All the Reactions That Will Work (arstechnica.com) 39
A team of researchers at Glasgow University have built a robot that uses machine learning to run and analyze its own chemical reaction. The system is able to figure out every reaction that's possible from a given set of starting materials. Ars Technica reports: Most of its parts are dispersed through a fume hood, which ensures safe ventilation of any products that somehow escape the system. The upper right is a collection of tanks containing starting materials and pumps that send them into one of six reaction chambers, which can be operated in parallel. The outcomes of these reactions can then be sent on for analysis. Pumps can feed samples into an IR spectrometer, a mass spectrometer, and a compact NMR machine -- the latter being the only bit of equipment that didn't fit in the fume hood. Collectively, these can create a fingerprint of the molecules that occupy a reaction chamber. By comparing this to the fingerprint of the starting materials, it's possible to determine whether a chemical reaction took place and infer some things about its products.
All of that is a substitute for a chemist's hands, but it doesn't replace the brains that evaluate potential reactions. That's where a machine-learning algorithm comes in. The system was given a set of 72 reactions with known products and used those to generate predictions of the outcomes of further reactions. From there, it started choosing reactions at random from the remaining list of options and determining whether they, too, produced products. By the time the algorithm had sampled 10 percent of the total possible reactions, it was able to predict the outcome of untested reactions with more than 80-percent accuracy. And, since the earlier reactions it tested were chosen at random, the system wasn't biased by human expectations of what reactions would or wouldn't work. The research has been published in the journal Nature.
All of that is a substitute for a chemist's hands, but it doesn't replace the brains that evaluate potential reactions. That's where a machine-learning algorithm comes in. The system was given a set of 72 reactions with known products and used those to generate predictions of the outcomes of further reactions. From there, it started choosing reactions at random from the remaining list of options and determining whether they, too, produced products. By the time the algorithm had sampled 10 percent of the total possible reactions, it was able to predict the outcome of untested reactions with more than 80-percent accuracy. And, since the earlier reactions it tested were chosen at random, the system wasn't biased by human expectations of what reactions would or wouldn't work. The research has been published in the journal Nature.
But can it write its own research papers? (Score:5, Interesting)
And equally important: can it be networked to similar machines (preferably made by other manufacturers and run by different labs) to set up its own peer reviews?
And how soon before the drug cartels are buying up every machine that is produced to discover new substances?
Re:But can it write its own research papers? (Score:5, Funny)
The more pertinent question, can it fake its own data for its research paper ;-?
Re: (Score:1)
No, but it can find reactions for revolutionary new battery tech and simultaneously complain about how it'll never work on slashdot.
Re: But can it write its own research papers? (Score:2)
Automated industrial design (Score:1)
I wonder if it's possbile to automatically design an industrial scale chemical process, an entire factory complex?
Say I want to produce x amount of this or that product. What raw materials do we need? Where should we build the factory? What would it cost? What are the options?
The system should of course take into account costs, externalities and regulations. Automatic procurement of commodity materials should be possible as well.
Re:Automated industrial design (Score:4, Interesting)
It would have to learn how to do the chemistry at factory scales to do that. This is possible but unlikely to be economic.
If anything, the interesting work would be to go in the other direction. It is possible to make lab-on-a-chip units where the reagents are sealed in and each experiment is done in a small drop. This means you can repeat each experiment many times and see whether the results are repeatable. If your process is catalysed by some low concentration impurity, then that effect will vary with the number of molecules of that impurity, and you would expect a larger scatter in the yield.
Interesting (Score:5, Insightful)
A carefully selected group of these working in parallel could theoretically parse the entire possible set of reactions, given sufficient time. (Yes, I know that with infinite molecular weight, there is an infinite number of possible compounds. However, only so many heavy molecules are interesting or useful, and of those, there will be certain classes that are more interesting than others. This approach would permit investigation of pathways without actually expending reagents, once its models are accurate enough. That means after a certain amount of training, a theoretical molecule of interest could be presented to the AI, and it could shit out the ideal synthesis pathway, and the next efficient arbitrary "n" pathways.)
This is the kind of thing that is the beginning of universal replicators.
Re: (Score:1)
and it could shit out the ideal synthesis pathway, and the next efficient arbitrary "n" pathways.)
Of all the places it could produce information, does it really need to come out of it's anus?
Re: (Score:1)
I think all compounds below a certain arbitrarily high molecular mass is a lot larger of a search space than you think. And if you want to try any possible experiment you really need one large machine more than a bunch of smaller machines however a matrix of smaller machine could fill in holes in the knowledge base for limited number of reactors. the number of possible reactions input combinations is 2^N where N is the number of known compounds, which will go up as the reactions uncover new compounds. I
Re: (Score:2)
somebody explain me (Score:2)
why would you use AI with a 80% success rate for this?
isn't chemistry all maths? can't you just calculate the same reactions and have a 100% success rate?
Re: (Score:2)
And, how does pattern-matching on 72 known products not just give you all the
products that are similar to the known set? How is that testing "all possible" combinations?
Comment removed (Score:5, Informative)
Re: (Score:1)
Simulating reactivity accurately requires absolutely stupid amounts of compute. Also FORTRAN: https://en.wikipedia.org/wiki/List_of_quantum_chemistry_and_solid-state_physics_software
Really "crude" analyses to predict absorption spectra of small biomolecules took everything we could throw at it in 2010, and we weren't exploring molecular interactions at all. The algorithms can be O(n^8) or worse.
Re: (Score:2)
How many really great super computer labs can a nation spare from its nuclear weapons simulations?
Because chemistry is HARD (Score:2)
>isn't chemistry all maths? can't you just calculate the same reactions and have a 100% success rate?
Because chemistry is applied quantum mechanics, and the math is HARD. I think they finally managed to simulate the chemical properties of a hydrogen molecule, but last I heard that was the most complicated molecule we've been able to simulate accurately - and it's literally the simplest molecule that exists, by a wide margin.
Inputs (Score:5, Interesting)
Why Patents are Absurd (Score:3)
This is a prime example of why patents are absurd and should be discontinued.
Re: (Score:2)
All too much work that results in patents is simply a matter of methodically working through combinations. I have invented a lot of things that way. Then I put them into manufacture and sell the products. That is the proper way to make money from your ideas. Patents ban other people from having similar ideas and using them. Patents are a bad idea and are being used to stifle innovation rather than the original purpose of promoting innovation.
Let me guess.. (Score:3)
The skeptic in me wonders if 80% of the combinations had no reaction
And when it's done ... (Score:3)
Re: (Score:2)
It uses an "Undo" feature to reverse the reactions until it gets back to its base components.
kind of interesting... (Score:2)
Re: (Score:2)
Point that AI ... (Score:2)