I am studying drug-protein interactions via Autodock Vina. I got my docking results but the problem is that the drug-target (enzyme) that I am studying has no 3D structure in PDB database. So I build one using SWISSMODEL with a template having 45% sequence identity (no other options). I had to add HEME group to my model via superimpostion using template PDB. I am worried about the accuracy of my protocols and results. Any suggestions on how to minimize errors in such cases??
Hi,
the best way is to get a ligand-protein complex concerning the protein you used as a template and analyze your system by comparison. Next, I would suggest a MD simulation to verify if your system is stable enough!
Regarding protein structure validation, observe ramachandran plot andnconsult literature likewise.
Regarding interaction studies, also do consut a lot of literarure and finally further validation will be added by MD simulation
Thirdly, try using a good simulation softwares. E.g. I would prefer schrodinger over any other simulation software and webservers.
Hi
In the absence of a co-crystal structure, the best way to validate docking is drawn through the correlation value of docking scores and bioactivities. Here, you must have a large ligand data set (n ≥ 36).
In case of any problem, refer to our articles:
(1) https://www.sciencedirect.com/science/article/abs/pii/S1734114018305723?via%3Dihub
(2) Article Toward resistance-compromised DHFR inhibitors part 1: Combin...
- Badges
- Science topic
- Similar topics
- Engineering
- Modeling