Should I sow QM in my fields?

Should I sow QM in my fields? was presented by Dr Ewa Chudyk, Senior Scientist, Evotec at the Cresset European User Group Meeting 2015.

Ewa compared Cresset’s molecular mechanics force field with some more computationally intensive QM approaches. Her work showed that both the QM and the MM approaches were proven to be complementary and useful in inhibitor optimization. She also concluded that the QSAR approach in Forge was equally as good as the QM method.

Evotec are a contract research organization. Ewa set out to assess how useful Cresset’s tools can be for Evotec’s drug design projects. Her group was interested in comparing Cresset approaches with QM approaches. Inspired by Cresset fields, she also investigated fields around the proteins and whether it is possible to observe complementarity between fields around proteins and ligands.

When trying to describe molecular systems simpler, faster approaches were taken. The MM Cresset approach is simple and fast, but it does not cover some effects that may be important for binding. To cover these effects they move to the QM space, but at a computational price.

In this talk Ewa evaluated whether the MM approach is effective enough for binding calculations.

At Evotec they use the FMO method. It divides the system into small fragments or residues. The interaction energies are calculated between each fragment or residue in terms of electrostatics, dispersion, exchange repulsion & charge transfer. They analyze what are the most dominant and most important for binding.

The presentation concentrated on one test case, CDK-2. This is an important enzyme because it is involved in the control of the cell cycle and its inhibitors might have anti-cancer properties. It is a good target for this study because there is plenty of QSAR available for it, plus an available set of 28 inhibitors. They carried out a structure decomposition on the most active inhibitors. The most active inhibitors aligned with high Cresset scores.

The FMO method can help to analyse the interactions around the inhibitors. The FMO analysis showed that different interactions dominate in every inhibitor part.

Another method that Evotec uses is Electrostatic Complementarity (EC). This lets the user distinguish areas in the molecule that are attractive and repulsive. They carried out an EC analysis on the most potent inhibitor and found that there was a lack of correlation between the EC score and the inhibition, suggesting that the electrostatics were not a driving force in the binding.

Ewa started to test different inhibitors to get a better picture of what was happening. Firstly, she found that there is some correlation between the EC score and the activity. The inhibitor core interactions were dominated by the electrostatics but she also found that the solvation component improves the correlation.

She then carried the same QSAR analysis with Forge. The best model used the most active ligand as a reference. She concluded that the QSAR Forge model is almost equally good as the equivalent model, and importantly, gave accurate predictions.

She concluded that the QSAR performance with FMO (QM) and Forge (Cresset) are equally good and that both approaches have been proven to be complementary and useful in inhibitor optimization.