Follow us on social

Latest Posts

Stay in Touch With Us

For Advertising, media partnerships, sponsorship, associations, and alliances, please connect to us below

Email
info@globaltechoutlook.com

Phone
+91 40 230 552 15

Address
540/6, 3rd Floor, Geetanjali Towers,
KPHB-6, Hyderabad 500072

Follow us on social

Globaltechoutlook

  /  Artificial Intelligence   /  IBM launches RoboRXN, a Chemistry Lab in Cloud, for Drug Discovery
IBM RoboRXN RXN Chemical machine learning drug discovery

IBM launches RoboRXN, a Chemistry Lab in Cloud, for Drug Discovery

Tech Giant IBM plans to speed up Chemical Research through RoboRXN

 

Technology giant IBM recently launched a chemistry laboratory called RoboRXN, in the cloud. This system leverages deep learning algorithms, IBM’s cloud, and robotic labs to automate the entire process and assist chemists in their work without requiring physical presence in a research lab. RoboRXN will be used to produce chemicals and drugs in a laboratory with little to no human intervention. Scientists can draw the skeletal structure of the molecular compounds they want to create. Plus, the system can be used to predict the materials required and plan the order in which they should be mixed. Project manager Teodoro Laino at IBM’s Zurich, Switzerland, and his colleagues used a dataset of chemical procedures built with their RXN neural network. They integrated it with cloud-based control of an automated synthesis machine.

According to IBM’s newsroom blog, the objective behind the RoboRXN platform is to dramatically speed up the drug discovery process by predicting the recipes for compounds and automating experiments. By speeding up this process, it is hoping to lower the costs of drug development and allow scientists to react faster to health crises like the current pandemic, where social distancing policies have slowed down lab work. Till now, the drug discovery cycle for new drugs and materials usually takes an average of 10 years and US$10 million to bring to market. And a greater share of this duration is spent on trial and error methods to synthesize new compounds. RoboRXN can assist the scientists by finding the most optimum scientific route and guides them to the most appropriate commercially available material to be started with. Once these data have been submitted, RoboRXN executes the process, and then sents the results to the scientists. Currently, Robo­RXN can manage up to five synthetic steps without human intervention.

In a live demo on 26 August, conducted from Zurich, Switzerland, RoboRXN chose a route for producing 3-bromobenzylamine, one of 3000 molecules the company had previously identified as potential COVID-19 therapeutic candidates. RoboRXN identified a synthetic route reducing 3-bromobenzonitrile using lithium aluminum hydride and produced instructions to tell a liquid-handling robot to do the preparation. In a demo, IBM researchers compared the process with the cooking of an apple pie. “Each component — such as the pastry — requires a specific set of instructions,” says Chris Sciacca, communications manager for IBM Research Europe. She continues, “You have ingredients like apples, sugar, flour, a binding agent, and so on. Then you follow instructions on how to blend these ingredients to make the final pie.”

The whole project started three years ago when IBM began developing machine learning models to predict chemical reactions. The neural network, which was trained on 2 million chemical reactions, was first introduced in a paper presented by the IBM Research team at the NIPS 2017 AI conference. The following year, IBM developed the AI into RXN for Chemistry, a cloud-based platform for chemical research, and presented it at the American Chemical Society annual exposition. Then in 2019, Laino’s team added a retrosynthesis tool to enable users to draw a molecule and have the software design synthesis for it. As per IBM, its RXN tool for retrosynthesis and reaction prediction outperforms all data-driven models on reaction prediction, picking the correct outcome with 90% accuracy.