Hi,
I'm trying to model the interaction between a protein and a glycopeptide (a hexose and around 8-10 aminoacids), but I don't have 3D structure information about this molecule (It was constructed de novo and I only know the sequence and the position of the hexose). I was thinking about use MD simulations in explicit solvent in order to have an idea about the conformation of my molecule. Do you know another approaches to predict the 3D structure?
Thank you.
You may perform MD folding simulations using an enhanced sampling technique such as REMD. You can try normal MD, but you may not be satisfied with it cause of the underlying rough energy landscapes of proteins. This means that you possibly won't be able to cross significant energy barriers with constant temperature MD and the folding will get stuck at some point. I'd also recommend using explicit water. We have recently shown that pure implicit solvent is not reliable for protein structure determination, but hybrid methods may compromise between accuracy and performance. Reduced computational expense can be achieved with REMD adaptions like our TIGER2h algorithm that utilizes an hybrid explicit/implicit solvent and thereby saves replicas.
Article Replica Based Protein Structure Sampling Methods: Compromisi...
A new study is submitted and may be soon accepted, that improves the accuracy of TIGER2h and eliminates most of the influence of the implicit solvent driven exchanges (TIGER2hs) by introducing explicit water shells for the hybrid approach.Generally, performing such folding simulation will require some effort and learning. For this length of the peptide however, it might be relatively easy to obtain satisfying results. TIGER2h or TIGER2hs may do it on a single GPU enabled computer within about 3 weeks. For geniune REMD or TIGER2A on the other hand, you may require access to a HPC system.
You can also use MD simulations for alone protein, alone ligand and protein-ligand complex with varying temperature and duration to get more insight towards the binding mode of complex.
Thank you so much for all your suggestions.
Mr. Norman Geist , efectively I was thinking about use replica exchange molecular dynamics in order to have more insight about the conformation of this molecule, but here in Colombia I don't have computational resources to do an in-depth folding simulation, I think that your suggestions about the use of the algorithm TIGER2h could be a solution because I can have GPU enabled computers.
Dr. Naresh Kumar , thank you for your suggestion, that could be a good approach in order to see the binding mode of this molecules, again I would need more computational resourcers, so I will try to get that.
Thank you both for your contribution.
- Badges
- Science topic