A fragment library representing available chemicals

We are interested in developing methods to help the discovery of new medicines or drugs. For most diseases or medical conditions, the aim is to identify chemical compounds that bind specifically to a particular protein or target, changing the way that protein works and thus affecting the disease. The drug discovery process begins with finding something that binds relatively weakly; small modifications are then made to this initial hit compound to improve its properties to be good enough to be a drug. The challenge is that it is difficult and sometimes impossible to find initial hit compounds for many protein targets by screening collections of large compounds. A new idea in drug discovery is to find (screen) small bits or fragments that bind and then either grow or merge these fragments to provide the hit compounds for optimisation. The aim of this project was to develop software methods to design the fragments used for screening so that, if a hit is found, then there is a higher chance that gical molecule and change the way it works - thus affecting the disease. The objective was to develop software that would design a small library (500 compounds) of low MW (<250 Da) compounds that would be maximally similar to the compounds of higher MW available commercially (typically many 100,000s of compounds). The purpose of such a library is to identify compounds that bind to a protein target - to provide starting points for new inhibitors. This would mean that when any fragments are experimentally found to bind to a particular protein, there would be larger, similar compounds available that could be tested for higher affinity binding.