DailyPost 2915
GOOGLE DEEPMIND – AI DESIGNED DRUGS
Founded in 2010, DeepMind can be called the original artificial intelligence company and was able to cross many barriers in the nascent field, in which the company and its founder Demis Hassabis had complete faith. In 2014 it joined the Google fold, getting in return both the astounding computational capability as well as deep pockets, which could keep DeepMind’s path breaking research moving without any hindrance. The development of AlphaGo and its comprehensive win over the Go world champion, proved that AI had arrived. DeepMind spearheaded the AI research and might be its inability to take the path breaking research to production, in hindsight seems to have become a handicap.
ChatGPT, and its parent company OpenAI stole the show, by its launch on 30th of Nov. 2022 and its global acclaim and acceptance. LLMs became a household name and tool. The IT giants got into the product mode with new ones appearing onto the scene with astounding regularity. When AI is redefining the world in entirety, there is enough for everyone to experiment upon and make a success out of it. Google DeepMind could not be left behind. Under an unified AI team, DeepMind has bounced back. DeepMind’s recent revelation has the capability to transform drug discovery and human health, totally.
What does DeepMind have to offer? In collaboration with Isomorphic Labs, Google’s DeepMind has predicted over 200 million protein structures using AlphaFold. AlphaFold was trained on nearly 100,000 known proteins. It has been driving significant breakthroughs in drug discovery targeting previously intractable biomedical challenges. From untreatable to treatable is the march, fascinating as never before in human history and that too with full precision. The model developed by this company can predict 3 structures of proteins. This would turn out to be the core kernel in driving significant breakthroughs in drug discovery.
An AI system named AlphaProteo was launched with the potential of generating novel proteins “designed to bind to specific target molecules”. We are thus at the cusp of significantly advancing research in drug design, disease understanding etc. But DeepMind is not the only player in the field. ESMFold Meta’s protein-folding model has also predicted about 772 million protein structures. Dreaming AI to bringing it to reality is what Hassabis is made of. For him the future is of AI designed drugs which address hundreds of debilitating diseases. Isomorphic Labs sees its future in becoming a multi-billion dollar business. Commercial potential of AI in drug discovery knows no bounds.
AI ENABLED DRUG DISCOVERY IS THE FUTURE OF HUMAN HEALTH.
Sanjay Sahay
Have a nice evening.