Interaction Profiling for Binding Site Comparison

Sebastian Salentin · Figshare · 2014

Natural compounds or drugs can be promiscuous, i.e. they may bind to several distinct proteins. Structural information on binding sites of protein-ligand complexes can be used to predict off-targets for ligands. These findings are especially useful for drug repositioning. While there exist many methods which use geometric or chemical features to judge binding site similarity, there has recently been much work on exploiting information on protein-ligand interaction patterns for this purpose. They consider the key binding features and can be easily encoded into fingerprints. Thus, they may unravel similarities even between remotely similar binding sites and are faster than methods based on structural alignments.

Read the paper · More papers on PaperTik