As the number of individuals who contract the Coronavirus (COVID-19) exceeds 97,000 and the deaths caused are already over 3,300, the scientific community is struggling to try to find a cure for SARS-CoV-2, the causative virus of the disease. DeepMind wants to help stop the outbreak and claims that AlphaFold, its artificial intelligence system, could be a key tool.
According to the British subsidiary company of Google, AlphaFold could be used to predict the structures of virus proteins through its algorithm. The process makes it possible to anticipate protein combinations when other similar examples are not available.
The structure of proteins is an important element in the process of understanding how they work, but the experiments needed to determine them can take months or even longer, says DeepMind. The company hopes to reduce the time it takes scientists to reach a discovery.
Given the current state of the COVID-19 outbreak, DeepMind indicates that it will make its forecasts available on an open license, even without having published its work in an academic magazine. The company's researchers warn of the fact that AlphaFold is still in development, so we cannot be completely sure that the results made available are accurate. However, the company confirms that the system was able to correctly predict one of the structures of SARS-CoV-2, which is already part of the Protein Data Bank.
Artificial intelligence has been used to improve the diagnosis of various diseases. DeepMind's AI system, for example, was able to detect breast cancers more accurately than medical teams. The technology has developed a model that has managed to reduce the number of false positive cases by 5.7% and false negative cases by 9.4%, during a test phase in some North American hospitals.
Recently, the American company C. Light Technologies developed a technology to help doctors diagnose multiple sclerosis more quickly. In the AI ​​system he developed, the goal is for patients to fix their eyes on a kind of target for ten seconds, while the technology analyzes the retina and predicts possible neurological disorders that the patient may have.
Also in February of this year, the British startup Exscientia and the Japanese pharmaceutical company Sumitomo Dainippon Pharma claim to have developed the first medicine created through an AI. DSP-1181 will be used to treat patients suffering from obsessive compulsive disorder.
