Molecular simulations: Docking of Ligands and Protein - Please bear with my complete ignorance... (Jan/30/2011 )
Hi all,
I am completely new in protein docking. I have to admit that I'm still learning (in fact I'm a slow learner...) and I totally got no previous experiences. I tried to learn from whatever manual I can get but still I do not understand....I hope somebody here got some experiences and can guide me.
I have a protein model, and I wish to do ligand docking simulation. Also, the ligand requires Calcium ions in order for it to binds on the active site.
My question is:
1) What is the best approach I can use to find the active site of the protein for the ligand binding.
2) Since the ligand requires calcium ions, should I dock the calcium ion first, or is there some settings I should have to configure.
3) How to analyze the result for the best ligand-docking? there are RMSD values, lower the better or higher?
Million thank you for your patience for read and answer my question.
For software, I do have:
PyMOL with autodock plugin
Chimera
Discovery Studio "Free Viewer" 2.5.5
MGL tools 1.5
Autodock 4, Autogrid and autodock vina.
Running windows 7 64bit OS
I do have a hard time in learn linux... thats the reason I use windows based software.
I really hope someone can help me here.
Million thanks again.
Adrian
The first thing to consider is the complexity of your problem. What are your molecules like? Or, more importantly what is your ligand like. If you have a rigid body complex, I would suggest the servers at rosetta commons. The drawback is you have to wait your turn in the queue, which can take some time (a couple of weeks in fact), but you wouldn't have to learn any programing and the results are more or less the best that you could get.
Now, the problem starts if your ligand is flexible. In that case the docking process is pretty useless unless you have a homologous complex with a known structure. If your ligand is a small molecule (which I'm guessing it isn't, but anyway) you could try Molegro Virtual Docker (MVD), which has a 30 day trial period and they are very keen to help you with your problem.
Now to answer your questions:
1. The active site of your receptor/protein is usually a pocket. So you could do a surface scan to check for cavities. For example MVD does this on its own and I must say it's a great feature.
2. The Ca2+ is probably only co-ordinating the ligand residues so they attain the correct position. Therefore "docking" of the calcium is futile, because it will tell you nothing about your ligand (besides the fact that you would be "docking" a single atom). However, it does influence the position of the ligand residues.
3. Most of the software packages have an energy function that is used to characterize the docking poses. The RMSD is used when comparing the output poses to an experimentally obtained structure. RMSD = root mean square deviation and is a measure of how well the computer derived and experiment derived models overlap. So, in short, the smaller the better.
I hope it helps. I think it would be a good idea for you to define why you are doing the docking and what you wish to gain from this information.
Also, linux is not a single OS, there are many distributions, some of which (for example Ubuntu) are very user friendly, have a GUI and are not difficult to learn to use. Another thing is to learn to use the docking software packages, which usually have a steep learning curve.
Best of luck.
Miha
BioMiha on Mon Jan 31 08:25:36 2011 said:
The first thing to consider is the complexity of your problem. What are your molecules like? Or, more importantly what is your ligand like. If you have a rigid body complex, I would suggest the servers at rosetta commons. The drawback is you have to wait your turn in the queue, which can take some time (a couple of weeks in fact), but you wouldn't have to learn any programing and the results are more or less the best that you could get.
Now, the problem starts if your ligand is flexible. In that case the docking process is pretty useless unless you have a homologous complex with a known structure. If your ligand is a small molecule (which I'm guessing it isn't, but anyway) you could try Molegro Virtual Docker (MVD), which has a 30 day trial period and they are very keen to help you with your problem.
Now to answer your questions:
1. The active site of your receptor/protein is usually a pocket. So you could do a surface scan to check for cavities. For example MVD does this on its own and I must say it's a great feature.
2. The Ca2+ is probably only co-ordinating the ligand residues so they attain the correct position. Therefore "docking" of the calcium is futile, because it will tell you nothing about your ligand (besides the fact that you would be "docking" a single atom). However, it does influence the position of the ligand residues.
3. Most of the software packages have an energy function that is used to characterize the docking poses. The RMSD is used when comparing the output poses to an experimentally obtained structure. RMSD = root mean square deviation and is a measure of how well the computer derived and experiment derived models overlap. So, in short, the smaller the better.
I hope it helps. I think it would be a good idea for you to define why you are doing the docking and what you wish to gain from this information.
Also, linux is not a single OS, there are many distributions, some of which (for example Ubuntu) are very user friendly, have a GUI and are not difficult to learn to use. Another thing is to learn to use the docking software packages, which usually have a steep learning curve.
Best of luck.
Miha
Hi Miha,
First of all thanks for your detail reply.
Actually I was trying to do docking between lectins and carbohydrate. I do have some hypothetical proteins which believed to be lectins and I was thinking of using molecular docking try to verify that based on the binding energy, at least this is what I was told could be done.
I will try to look into MVD.
Thanks again for your kind reply. If you do have any advice about docking a lectin you are very welcome.
Million thanks again.
Adrian
I did see a paper a while ago about docking a carbohydrate into the binding site of a Fab fragment using MVD. Don't remember exactly who the authors were, but you can Google e.g. "Fab AND carbohydrate AND molegro" and see what you get.
The Materials and methods section is pretty well detailed and MVD has a great Manual.
Best of luck.
Miha