The order of my ligand preparation is as follows:
1. I download the the ligand structure in Mol2 format from ZİNC database
2. I adjust its pH state via LigPrep pocket of Schroedinger
3. I use the search conformation of the MocroModel of Schroedinger to get the low energy conformations of the ligand by Monet Carlo method.
4. I employ 20 lowest energy conformation for my docking.
Is this process correct for preparing my ligands?
I would say that your procedure is reasonable. Correct implies that the constructed model and the performed simulations predict or reproduce experimental observables. The model and simulation accuracy is something that has yet to be validated or proved through data analysis. At the end of the day, one can construct several models with a variety of different scales (e.g. atomistic, united atom, coarse-grained), whose validity is shown by comparison to experimental data. However, there are well established methodologies that are so widely used and validated that researchers trust them to be predictive in nature without personally testing them.
I am not too familiar with the Schroedinger software itself, but here are some general thoughts concerning model construction and simulation execution. If you haven’t done so yet, I suggest reading papers who have implemented similar methodology using the Schroedinger software and comparing your workflow to theirs. Some issue to consider in your evaluation are a) how reliable is the downloaded ZINC 3D structure of the compound, b) could enantiomers of the compound exist, c) are the predicted protonation states reasonable, d) how reliable is MacroModel at predicting low energy structures, e) if the workflow includes partial atomic charges generation for the ligand, are they balanced with those of the receptor site when docking, and f) are compound and receptor flexibility taken into account in some way?
- Badges
- Science topic
- Similar topics
- Mathematics