The neural network has learned to create more than 300 drugs from chemical waste

Scientists have been working on using computer systems to create closed loop chemical production.

cycle.Using the Allchemy platform, the researchers built giant synthetic interconnect networks. The proposed solutions make it possible to reuse more than 200 types of chemical waste and produce about 300 types of drugs, such as carvedilol and salbutamol.

Chemical industry, includingpharmaceuticals, cosmetics or fertilizers, create a large amount of waste, consisting of many complex chemical molecules, the researchers note. Scientists have long been looking for ways to use waste from one industry to create other products. But, as the authors of the new work note, this was extremely difficult to achieve, since each compound opens up a million paths to other chemical formulas. Given the number of factors that must be taken into account in the calculations, the analysis becomes too difficult.

In their work, scientists used computerthe Allchemy system, which can analyze the composition of complex objects and, on its basis, build possible synthetic networks (options for the synthesis of new substances). Initially, this system was used to look for chemical signs of life on other planets.

Platform AI forms a network of synthesis routes,corresponding to all inputs, and then selects those that correspond to circular green production and ranks them according to given criteria.

A small section of the network of possible routes. Source: Agnieszka Wołos et al., Nature

The researchers note that the received routesare descriptions of the steps involved in synthesizing the desired compound. The scientists successfully tested their system using several routes that were created to synthesize real drugs.

Chemists believe that the widespread introduction of computer systems will significantly accelerate the transition to closed-cycle chemical production and reduce the amount of toxic waste that needs to be disposed of.

Read more:

Inexplicable duality found in elementary particle physics: what will it lead to

Look at the alignment of the planets in 2022: the brightest planets are visible in the photo from Earth

Scientists explain how a stem cell "understands" what it will become