New drugs to keep us healthy will come to market faster thanks to a new smart system developed by the Technion-Israel Institute of Technology. The smart system will be presented this week at the smart system conference taking place in London.
Built on a construct of artificial intelligence (AI) and deep learning, the smart system will significantly shorten the time it takes to bring a new drug to market. It will also reduce costs and these production savings can be passed on to consumers.
Right now, bringing a new drug to market takes 10-15 years and costs between $500 million to $2,5 billion dollars! Sounds unbelievable, right? In fact, these numbers are very common in the world of drug development.
Many years ago, a drug that actually worked, was usually discovered by accident, for example, penicillin. However, in the modern era with advanced technology and high speed computers, new drug discovery has been comparatively faster. Nevertheless, it’s just not fast enough.
Many diseases today, do not have an effective long-term drug treatment protocol. For example, Alzheimer’s disease, dementia, and Parkinson’s disease. This failure, takes a toll on patients, their families and the economy.
The problem is, there are billions of molecules that show therapeutic potential. Which ones do you choose to test? Which will or won’t work? With the FDA stringent drug approval standards, drug companies certainly don’t have enough time or money to find out.
New Drugs: How The Smart System Speeds Up Drug Development
The Technion researchers came up with a new approach for this generation of therapeutic molecule candidates.
Their working hypothesis is that the drug development chemistry vocabulary is similar to that of natural language. The drug “language” was developed based on hundreds of thousands of molecules. They added the chemical composition of all drugs FDA approved up until 1950.
These approved drugs served as the prototypes to generate new potential drugs. AI and deep learning then generated many variations of the existing drugs.
In short, the smart system is based on a pharmacological language, data on existing drugs, and a creativity promoting mechanism.
When the researchers instructed the system to propose 1000 drugs based upon old drugs, they were surprised to discover that 35 of the new drugs generated by the system are existing, FDA-approved drugs. In other words, the investigators demonstrated the system’s efficiency in developing ‘rational’ or valid drugs.
New Drugs: Ramifications of AI’s Effect On Future Drug Development
Under the current system and schedule of drug development, the number of new drugs approved by the FDA declines at a rate of approximately 50% every 9 years. This is based on the ratio between the number of new drugs to the investment in research and development. Bringing fewer drugs to market will have a tremendous negative impact on health care.
According to the Technion researchers, their AI algorithm targets the creative stage of drug development at the molecular discovery stage.
This allows the computer to evaluate huge amounts of data and pick out new molecule connections based on a previously approved prototype. The computer identifies potential successful molecule candidates.
This AI, deep learning paradigm will speed up the approval of effective drugs while dramatically reducing costs for drug manufacturers and consumers.
A win-win strategy for all concerned and a boon for the state of health care.