A team of scientists from the University of Sheffield have developed a method for assessing whether protein structures are accurate.
The method can be applied to computer predictions, and to experimental structures from nuclear magnetic resonance (NMR) studies, and could pave the way for the design and development of new drugs.
Proteins are the workhorses of the body and the targets of most drugs, which are designed to bind to specific regions of proteins. For drug development, you need an accurate structure of the protein. You also need some confidence in the accuracy, which is what this method provides. The method will therefore be very useful for drug development, because it will identify the best starting points.
The method uses nuclear magnetic resonance (NMR) to obtain experimental measurements on what the protein looks like in solution, and compares these to the structure. NMR works in a similar way to an MRI scan but on a molecular level.
There are three main methods for determining protein structures. One is X-ray crystallography, which already has good methods for determining their accuracy. One is computer predictions based on artificial intelligence, which have recently shown impressive improvements and are likely to grow rapidly in importance because they are so quick. The third is NMR.
A longstanding problem with NMR is that, until now, there has been no good way of analysing how accurate the models are. The new method looks set to solve that problem. There is currently a database of protein structures which contains over 13,000 NMR structures. Many of these could be used for drug discovery but because there has previously been no way to check their accuracy they have not been used effectively.
The team of researchers at Sheffield will use their new method to analyse the accuracy of some of these structures, which will open up many possibilities for drug design. The team have already applied the method to over 6,000 of these structures. Their work has been published the journal Nature Communications and features in a collection of Editor's Highlights, comprising recent articles across structural biology, biochemistry and biophysics.
Professor Mike Williamson, lead author of the research from the University of Sheffield, said: “We are very excited by this work, which solves a problem that has been apparent since the first NMR structure in 1985. We are expecting it to speed up structure determination and drug design very significantly.”
