Have you ever wondered how your doctor decides what drugs to prescribe to treat your ailments? It may be likely that there may be previous references, and the training received provides some knowledge, but it is common knowledge in the medical community, that all drugs do not react in the same manner in all humans. In cases, such as the recent Ebola outbreak that originated in West Africa, there was no cure, and with traditional testing methods, it would have taken years to shortlist effective treatments, but the new Startup, known as Atomwise, has apparently changed the face of the research for finding and developing better treatments. With the use of artificial intelligence algorithm, similar to Google’s Deep Learning, it took Atomwise four months to identify two medications with potential stop the spread of the virus, and to move testing into clinical trials.
New molecules for treatments can cost close to $2 billion and may take up to 15 years to be brought to market, but with the power of computing and machine learning, Atomwise appears to be on the verge of ushering a new paradigm in research.
CEO Abraham Heifets, explains that Atomwise, which was born in the incubator program of the University of Toronto, has developed an machine algorithm that exploits much of the the new techniques in machine learning and utilizes artificial intelligence process flows to make faster and more accurate predictions. The computers are used to analyze millions of different medications, and determine how likely they are to counteract the signatures of various diseases with minimal side effects. The advantages are that the medications are already in existence, but in addition to discovery, the Atomwise machine learning algorithm may also help in development of new treatments.
Dr Heifets and his small team is hopeful that cures for many current diseases, which are believed to be incurable, may actually already be on the shelf, and we just don’t not know it. In the same manner that acetylsalicylic acid, better know as Aspirin can be effective in preventing cardiac disease, Atomwise is currently using its machine learning algorithm to investigate the potential treatments for several other chronic diseases.
The use of predictive models in the pharmaceutical industry is not really new, but Atomwise has taken it to a new level, with what is called deep learning. Much of the guesswork or human analysis is removed, and it is the computer, which is taught to mimic the human thought process is used to make the decision. The results with the use of sophisticated algorithms are much faster decisions, saving of expensive resources, a higher degree of accuracy, and removal of factors due to human errors.
The company which has recently relocated to Silicon Valley in California, has attracted the backing of heavy venture caps, that include Khosla Ventures, Data Collective, which led the funding, Draper Fisher Jurvetson venture cap firm, AME Cloud Ventures, managed by Jerry Yang, the founder of Yahoo, and Byran Johnson’s OS Fund. Johnson was co-founder of Braintree, which was eventually sold to eBay for $800 million.
The company has recently raised more than $6 million in its latest round, and is working on several projects with companies such as Merck , and searching for treatments for conditions such as multiple sclerosis, Malaria, and brain cancer.