Dr. Sherman works closely with researchers using Schrödinger software for molecular modeling and drug design projects. In this newsletter, Dr. Sherman talks about displaying hydrophobic interactions, and more. All of the scripts discussed here may be downloaded free of charge from the Schrödinger Script Center, or by using 'Update...' from the Scripts menu in Maestro.
Q: I have just built a homology model and would like to automatically detect all disallowed phi/psi backbone or chi1/chi2 side chain torsional angles in my structure. Is there a way to do this?
A: Yes, you can do so using the Disallowed Backbone/Sidechains script (select_disallowed.py), which appears under the Workspace Tools category of the Scripts menu. This will detect disallowed phi/psi (backbone) or chi1/chi2 (side chain) angles in your Workspace structure. These residues will be selected in the Workspace so they can easily be corrected by running a Prime side chain prediction or loop refinement on the necessary residues. Additionally, the Ramachandran Plot under the Maestro Tools menu allows for interactive investigation of phi/psi values.
Q: I have tried to use the Contacts tab in the Measurements panel to display hydrophobic contact but it appears to show all contacts (not just hydrophobic). Is there a way to display only good hydrophobic contacts between a ligand and protein in the Workspace?
A: This can be done with the Display Hydrophobic Interactions script (display_hydrophobic_interactions.py), which is found in the Workspace Tools category of the Scripts menu. This script will automatically detect the hydrophobic contacts between the ligand and protein in the Workspace, and will display these contacts with green dotted lines.
Q: I read about Spectral Clustering as described in J. Chem. Inf. Model, 2007, 47, 1727-1733, and like many of the features of this method. Has Schrödinger implemented such a clustering technique?
A: Yes, we have implemented Spectral Clustering using our Canvas cheminformatics infrastructure (spectral_cluster.py). It can be found under the Cheminformatics category of the Scripts menu. You can select from a range of fingerprint types and similarity metrics. The cluster membership, cluster eigenvalue, and normalized cluster eigenvector will be added to the Project Table for each entry. The eigenvalues represent the overall intermolecular similarity within each cluster. The eigenvectors show the relative contributions from each molecule within a cluster.