The computer will calculate the side effects of the drugs in advance
A mathematical model has been created to predict the side effects of drugsCopper news based on the materials of Science News: Network analysis predicts drug side effects
American scientists have created a mathematical model that is able to identify the side effects of drugs before their therapeutic use. The study was conducted by a group of specialists from Harvard University Medical School in Boston under the leadership of Aurel Cami. A report on their work (Predicting Adverse Drug Events Using Pharmacological Network Models) is published in the journal Science Translational Medicine.
Kami and his colleagues analyzed information about drugs from the corresponding reference book dating back to 2005. Scientists have created a network that describes the links between the use of more than eight hundred drugs in clinical practice and the development of more than 850 types of adverse reactions to these drugs.
Using the data obtained, the researchers developed a new mathematical model. They "trained" her to identify similar relationships for new drugs for which information about side effects had not yet been accumulated.
As a result, the new model was able to predict the presence or absence of a specific side effect in the studied drugs in 42 percent of cases compared with the control. The control was carried out by "traditional" long-term monitoring of patients taking these medications.
Among the predicted side effects were tendon ruptures after taking the antibiotic norfloxacin, as well as suicidal thoughts in patients who were prescribed the antiepileptic drug zonisamide.
At the same time, Kami noted that the accuracy of the mathematical forecast varied depending on the type of drug. In particular, according to him, the presence of side effects was more difficult to predict for drugs used in the treatment of skin diseases.
In November 2009, a similar study was conducted by specialists from the University of California, San Francisco. With the help of computer analysis, they were able to predict more than four thousand unexpected potential mechanisms of action of drugs, both beneficial and undesirable.
Portal "Eternal youth" http://vechnayamolodost.ru
23.12.2011