News Release 

MIT Technology Review selects AI molecular design as a breakthrough and highlights Insilico

InSilico Medicine

IMAGE

IMAGE: MIT Technology Review highlights Insilico Medicine view more 

Credit: Insilico

26th of February, 2020, 7:00 AM ET - Insilico Medicine today announced that MIT Technology Review selected AI-discovered molecules as a breakthrough of the year with the availability horizon of 3-5 years. Insilico Medicine's research together with the University of Toronto was highlighted.

'The ability of deep learning and other AI tools to find novel molecules with desirable properties will transform drug discovery. It promises to make the development of new medicines far faster and more effective, and is an important new tool in the hunt for better drugs. Insilico Medicine has been a leader in using some of the most exciting techniques in AI, such as GANs, for drug discovery. And, with the latest results, GANs is proving to be a powerful new tool for finding promising molecules.' said David Rotman, Editor at Large.

On the same day, Insilico Medicine announced the launch of its Entrepreneur in Residence program in brain cancers with the objective to externalize the effort before August, 2020. The program will be led by pharma industry expert and seasoned R&D scientist, Dr. George Okafo.

'I remember the first AI and drug discovery conferences in early 2016 when we presented the first theoretical basis for using the Adversarial Autoencoders (AAE) for the generation of new molecular structures with desired properties. I had to spend a lot of time explaining the principles using images and the deep learning folks did not get the chemistry part and chemists were lost in math and asked for experimental validation. No one took it seriously. Now, most pharmaceutical companies started their internal generative chemistry groups. I am very happy to see that', said Alex Zhavoronkov, PhD, co-founder, and CEO of Insilico Medicine.

Generative AI is an emerging technology in both chemistry and biology. Insilico Medicine is working in these areas since 2015 and holds the critical intellectual property. According to the recent opinions, drugs generated using AI may be classified as being invented by AI. Insilico published its first peer-reviewed papers in this area in 2016, multiple theoretical papers in 2017, and recently published several proofs of concept for JAK3 and DDR1 kinases with experimental validation. It recently generated and released a range of molecules targeting COVID-19 protease generated in 4 days. Generative biology technology for the target discovery is developed and utilized internally since 2015 and the first landmark experiments are expected to be published in 2020.

'Drug discovery is a special industry where secrecy is encouraged and many advances cannot be openly shared. Generative chemistry is in a more advanced state than what can be judged from published work as most of the efforts in difficult targets are internal. I hope that within the next couple of years it will be possible to publish these advances, and the pharmaceutical companies will also publish their achievements, and molecules designed using generative chemistry will progress into the clinic', said Alex Aliper, PhD, co-founder and president of Insilico Medicine.

###

About Insilico Medicine

Insilico Medicine is an artificial intelligence company with offices in six countries and regions striving to accelerate three areas of drug discovery and development: disease target identification, generation of novel molecules (generative chemistry) and synthetic biological data (generative biology), and prediction of clinical trial outcomes. The Company was the first to apply the generative adversarial networks (GANs) and reinforcement learning (RL) to generate new molecular structures with the specified parameters in 2015. In addition to collaborating with large pharmaceutical companies, Insilico Medicine is also pursuing internal drug discovery programs in different disease areas.

Website: http://insilico.com/.

Disclaimer: AAAS and EurekAlert! are not responsible for the accuracy of news releases posted to EurekAlert! by contributing institutions or for the use of any information through the EurekAlert system.