February 27, 2012
Text: Elke Maier
In order to find a new substance that may one day find its way into a new cancer drug, scientists must first detect suitable drug candidates among all of the possible molecules. Once they have discovered a promising molecule, they must synthesise it and test its effectiveness before its further development into a drug can begin – all in all a laborious and, above all, time-consuming undertaking.
In the early 1990s, researchers pinned their hopes on the assembling of as many molecules as possible with the help of synthesis robots, and then trawling through these vast substance libraries for suitable compounds with the help of automated procedures. However, the initial enthusiasm for this approach soon turned into disillusionment: very few of the newly discovered molecules displayed any effect on living cells, and almost none of them completed the process of development into a finished product.
As a result, Herbert Waldmann and his colleagues at the Max Planck Institute of Molecular Physiology in Dortmund are looking for more efficient methods. The researchers’ main concern here is to both limit the search and simplify the associated synthesis process. “There are simply far too many different compounds to search randomly on the off chance that you’ll stumble onto something,” says Herbert Waldmann. The chemical structural space that includes all possible drug-like structures contains an estimated 1062 different molecules – a number, which, written out in full, fills two thirds of a line on a closely spaced typed page. Therefore, pre-selection is the most important thing.
To this end, the scientists search the chemical structural space using an ingenious computer program called Scaffold Hunter, which was developed at the Max Planck Institute in Dortmund. Scaffold Hunter generates maps of a selected chemical structural space based on structural criteria and enables the researchers to navigate the sea of possible molecules and approach islands of biological activity on the computer screen. “It actually works as simply as a video game,” says Herbert Waldmann. When navigating, the Scaffold Hunter searches for structural motifs that resemble already known structures with a particular biological characteristic. Because chemically related compounds are also very likely to have similar properties, the researchers can track down promising structures in this way. They can then use these as a basis and experiment with different chemical appendages to synthesise new compounds.