Identifying Protein Binding Hot Spots by Using Deeply Buried Atomic Contacts

Solvent accessible surface area (SASA) is an important descriptor of protein residues and atoms. It is widely used as an outstanding feature in hot spot prediction by many computational methods. However, SASA is not capable of distinguishing slightly buried residues, of which most are non hot spots, and deeply buried ones that are usually inside a hot spot. In this work, we propose a new descriptor for residues, atoms and for atomic contacts, namely “burial level”, which can capture the depth the residues are buried. We identified the number of different kinds of the deeply buried atomic contacts at different burial level that are directly broken in alanine substitution, and we used these values as input for SVM to classify between hot spot or non hot spot residues. We got an F measure of 0.6237 under the leave-one-out cross-validation on a data set containing 258 mutations. This performance is better than other computational methods.