Side effects can be assessed before the development of drugs
Computer model successfully predicts side effects of drugs
The principle of its operation is to compare the molecules that make up the drugs with a list of molecules that can bind to 73 target proteins of the human body and cause serious side effects, for example, a heart attack. A description of this model can be found in the journal Nature (Lounkine et al., Large-scale prediction and testing of drug activity on side-effect targets).
A group of researchers from the School of Pharmacology at the University of California (UCSF School of Pharmacy), USA, Novartis Biomedical Research Institutes and SeaChange Pharmaceuticals, Inc. are currently testing a new model on modern drugs whose side effects are well known and studied. To date, the model has predicted the presence of 1,241 possible side effects in 656 drugs, 348 of which matched the Novartis database. Another 151 side effects not previously established for these drugs were later confirmed during tests. According to scientists, no modern technique can predict the presence of potential harmful effects with such accuracy.
The presence of serious side effects remains the second most common reason why a new drug is rejected (inefficiency remains the main reason). As a rule, the development of a new drug takes up to 15 years and about 1 billion US dollars, and according to recent data — from 4 to 12 billion. If, as a result, the drug has an important side effect, this money and time will be wasted.
In fact, these billions include the cost of developing 10 drugs, 9 of which do not pass one of the stages of research. Sometimes — the last phase or even the stage of post—marketing control, when it is necessary to recall a drug already released on the market - VM.
"The new computer model can serve as a "safety cushion" in this case," explains Michael Keiser from the University of California. "When you decide which of the tens of thousands of components to combine to produce a new drug, the model will tell you which ingredients will be the safest." The model can also help to review existing drugs on the pharmaceutical market and find hitherto unknown side effects. For example, it has long been known that a synthetic analogue of estrogen can cause abdominal pain, but the reason for this has so far remained unknown. Computer analysis showed that this drug binds to the target enzyme cyclooxygenase type 1 (COX-1). Nonsteroidal anti-inflammatory drugs, in particular, acetylsalicylic acid, which also causes abdominal pain, the formation of stomach ulcers and gastric bleeding, are also associated with the same target.
Portal "Eternal youth" http://vechnayamolodost.ru14.06.2012