Release date: 2017-08-15
Developing drugs is a very expensive and time consuming task, and failure of drug trials can cause millions of stocks to plummet. Therefore, the sooner you identify the potential to develop into a viable drug, the better.
As such, pharmaceutical companies like GlaxoSmithKline (GSK), Merck, Sanofi and Johnson & Johnson are turning to artificial intelligence (AI) to help them.
Professor Andrew Hopkins is the CEO of Exscientia, an AI-based drug discovery company that recently signed a £33 million deal with GSK. He said that AI and humans working together in the so-called "half-horse team" can help save three-quarters of the time and cost. In Greek mythology, the Centaur is half-human and half-horse, and its strength is very strong and the movement is fast. Professor Hopkins believes that AI gives scientists more power.
Pamela Spence, head of the global life sciences industry at consulting firm EY, said successful drug discovery relies on an accurate understanding of the effects of the disease. “Once you know, scientists will search for molecules that can selectively interact with this 'target' and reverse the process or slow down the damage – the key is to 'hit' the molecule,†she explained.
Scientists often target a disease as a weapon to solve this problem. The traditional method of drug discovery is that a small team of scientists carefully tests each potential target and hits the molecule, hoping to find the ultimate goal - but this drug discovery process is a very time consuming method with a very high failure rate.
Therefore, the introduction of AI is like having a research assistant who can solve problems at an incredible speed through system search. She added: "You can first identify what might work through the AI ​​supercomputer and determine what doesn't work." This is the medical term for computer research, not "in vitro" - test tubes and "in vivo" - Testing of animals and humans.
Since performing human clinical trials accounts for the vast majority of drug discovery costs, the sooner we can set targets, we can save costs. Ms. Spence said: "Physical testing can be done with a smaller number of potential new drugs and can ensure a higher success rate."
Exscientia's AI algorithm contains a large amount of data, including the structure of the disease, the efficacy of existing drugs, peer-reviewed studies, and microscopic sample observations. Professor Hopkins said that all these possibilities are shrinking step by step. "We are not going to rule out uncertainty, because it is a messy and cumbersome data, and there is a very interesting analogy between human creativity and evolution."
Professor Hopkins added: "Our goal is to propose small molecules as candidates for up to 10 disease-related goals, which can then be tested by clinical tests. The resulting pills are validated by precision tests..."
GSK recently established a research unit focused on enhancing drug research through the use of "computer" technology (including AI, machine learning and deep learning). The team is led by John Baldoni, head of R&D at GSK. He said: "We currently have many plans to be carried out. The cost of the entire drug discovery is about $1.7 billion (about 1.3 billion pounds). The cost from the laboratory to the clinic is about 33% of the above figure, which takes about five years. Half the time. Our goal is to reduce time to one year and reduce the cost commensurate."
AI is also exploring other ways in the drug discovery process.
For example, AI uses natural language processing to screen emerging literature, such as chemical libraries, medical databases, and scientific papers, to arrive at conclusions about possible new drug candidates.
Earlier this year, using AI technology, one of the drugs for the treatment of motor neuron disease (also known as ALS amyotrophic lateral sclerosis) was discovered, which prevented the death of motoneurons in the brain of patients. And delay the onset of disease in animals. Ken Mulvaney, founder and chairman of Benevolent AI, said: "We are very encouraged by these findings.
Patients should also be encouraged by these findings. AI-based drug discovery is expected to bring more effective and cheaper drugs to the market faster.
Source: billion euro network
Rice Seed,Paddy Seeds,Black Rice Paddy Seeds,Upland Rice Seed
XIKE AGRICULTURAL GROUP CO . .LTD. , https://www.laoseed.com