Final comments on the OppA system

No strong correlations have so far been observed between the experimentally and computationally determined energies of binding in any of the many computational procedures examined. There are a number of possible reasons for this. The first is that the OppA system has evolved as a relatively indiscriminate binding protein. The various amino acid side-chains that line the ligand binding pockets, and the solvent in them, avoid strong interactions with the ligand, only the main-chain is bonded tightly. Scoring functions such as LUDI, however, are parameterised on other systems where the geometry of interaction is optimal. It is also a challenge for the calculations to take into account the explicit disorder that is seen in solvent positions in some of the complexes.

Existing scoring functions reward for good interactions, but do not penalise for bad - particularly perturbations that occur in the receptor protein. Current work is focussed on scoring functions that take into account poorer contacts - the ongoing experimental programme of abnormal amino acids is challenging the side chain pockets and will provide additional data on the effect of these unfavourable interactions. In addition, the buried water molecules and flexibility of the ligands are not typical of most protein-ligand complexes.

We continue our search for other suitable experimental systems that will provide complementary data on a quite different system. Our experiences to date provide an extremely valuable set of criteria on which to choose such a system.

0 0

Post a comment