Molecular Modelling to Help Create Better, Safer Drugs
News May 24, 2013
The research should ultimately help predict how new drugs will be metabolized in the body, potentially helping avoid adverse drug reactions in future.
Professor Adrian Mulholland of the School of Chemistry and colleagues used molecular modelling to show in atomic detail how ibuprofen, diclofenac and warfarin are broken down by a group of enzymes called cytochrome P450s which play an important part in the metabolism of drugs.
Cytochrome P450s break down drugs by adding oxygen atoms to them, thus making them more soluble in water and easier to remove from the body. It's important that drugs are broken down in this way so they don't accumulate to toxic levels. However, it's also important that the drugs aren't broken down too quickly otherwise they won't stay in the body long enough to work.
Different people have different types of P450 which mean they break down drugs more quickly or more slowly. Potentially harmful complications can also sometimes occur, for example, other drugs can 'block up' P450s thus interfering with the metabolism of a particular drug. Other substances can also interfere with the process, for example grapefruit and grapefruit juice contain a molecule that 'inhibits' some cytochrome P450s, preventing them from breaking down drugs. This can cause the drug to build up to a toxic – and possibly lethal – level.
Professor Mulholland said: "An important aim in developing a safe, effective drug is understanding how it will be broken down in the body. This process would be made quicker, cheaper and safer if we could predict reliably – for example, by using computers – how a candidate drug reacts in the body.
"This study uses molecular modelling methods which are able to describe chemical reactions in large and complex enzymes such as cytochrome P450s. Our results agree well with experiments, and point to how modelling of this sort can help in developing predictions of drug metabolism."
An artificial intelligence (AI) approach based on deep learning convolutional neural network (CNN) could identify nuanced mammographic imaging features specific for recalled but benign (false-positive) mammograms and distinguish such mammograms from those identified as malignant or negative.
2nd International Conference on Biological & Pharmaceutical Sciences
Jul 12 - Jul 13, 2019
2nd International Conference on Computational Biology and Bioinformatics
May 17 - May 18, 2019