Thursday, May 17, 2007

The Art of FREE Molecular Docking (Using DOCK 6.1) (Part 2)

Step by step docking

In this article, I am only going to SUPPLEMENT the tutorial provided at the DOCK website. Mainly the things which I found troublesome for me. So you should check out the tutorial first for a better understanding on the steps.

The steps in docking using DOCK are:

1) Structure preparation (enzyme and substrate)
For structure preparation, we will have to modify the 3D protein pdb file we have. I used a modeling program called Chimera (as what the tutorial advised), which is available free of charge from the Chimera website ( Structure preparation was a straight forward process. I mean, you only have to follow the tutorial to do so (it's already quite clear).

Just remember to prepare the charged mol2 structure of the ligand bound to the pdb protein's active site. You will need it if you want to be more efficient in the grid and sphere generating process. Even though you are not going to dock the ligand. I didn't identify this from the beginning, which created a problem in the sphgen - sphere generating procedure (it took too long, it could take for about 8 hours on my 1.6 Ghz centrino laptop with 512 RAM).

I did the structure preparations on windows, since I wasn't successful in installing the Chimera program on openSuse 10.2. Still remains a mystery for me up to now.

When you're finished with the preparation, now we can continue to the next step.

2) Sphere generation and selection
For generating the molecular surface of the protein, you will need a program called dms (as I have mentioned in my previous post). Installation was quite simple. Just type “make install” in the dms directory and stop there. You should have dms installed on your system (that's what happened to me in openSuse anyway). Remember that this is installation for a local machine. For parallel machines, there are other things you need to modify (check the dms readme that comes with the package).

When finished generating the molecular surface, use the program sphgen to generate the spheres.

Tip 1: copy the programs you want to use in your working directory!!.
Tip 2: in openSuse, instead of just typing “sphgen (options)”, I had to type “./sphgen (options)” to get it done!

After finished generating the spheres, now it's time to select the spheres you want to use as the active site. You can use the largest cluster created by sphgen, use a sphere in a selected radius, or select the spheres manually. I would recommend the second choice (use a sphere in a selected radius). Here, you should use the mol2 structure of the ligand bound to the pdb protein's active site (mentioned in step 1). This will save you precious time, believe me!

3) Grid generation
For this part, I had no problems at all. I just followed the tutorial, and I was home free.

There are 2 main ways of docking. First is the flexible ligand docking, and second is the rigid ligand docking. The flexible docking lets dock change the ligand conformation in response to the binding site. The rigid ligand doesn't let dock change ligand conformation. You should choose the one which suits your need.

For the docking process, thankfully I had no problems (I used flexible ligand docking). However, you might want to check the parameters that you need to tamper. For docking, I followed the dock v5.2.0 tutorial which is also available at the DOCK website.

Interpreting results
There are two main results that I found interesting (well, up to now, these are the only things I can analyze =)). First is the interaction energy of the ligand-protein complex. The lower the interaction energy, the stronger the bond is. I found this as additional data to supplement my kinetic experiments. Second is the conformation and orientation of the substrates in the binding site. Comparison with established ligand-protein structures can give better understanding on the effect of binding site structure has on substrate binding.

Well, this is a little taste of what DOCK can do for you. I have still much to learn about docking. But, since I am quite sure that it will be very important in future enzymatic studies, I have to spend more time to study it.

The Art of FREE Molecular Docking (Using DOCK 6.1) (Part 1)

Introduction and Installation

Enzymes play an important role in living organisms. They catalyze chemical reactions, where a substrate is converted into a product. Substrates are molecules that bind into the enzyme's active site where it undergoes conversion into the enzymatic product. I am not going to go too deep regarding enzymes, since you can find similar resources for this in the Internet (Wikipedia is a good place to start off).

What I am going to talk about, is a pragmatic way to better understand the interaction between an enzyme and its substrate using molecular modeling, in this case, docking using DOCK 6.1 (check the DOCK website at: Well, mostly, I will be talking about my experience using the program, from installation to the docking process.

So what is docking?
Substrates, as do all molecules, have many different conformations they can take. The conformations play a determining factor in the way it binds to the active site of an enzyme. What also needs to be elucidated is the binding orientation, interaction energy (Ei), etc. One way to predict the orientation, conformation, and interaction energy (Ei) of an enzyme-substrate complex, is by molecular docking using in silico (computer) analysis. In short words, docking is meant to see the interaction between molecules and ligands (enzyme or protein). In my case, the data retrieved will then be analyzed to see substrate preference of the enzyme tested.

There is a wide range of programs that can be used for molecular docking, from freeware to expensive commercial wares. Among all, GOLD and Glide are considered to have the best accuracy. But to obtain one of them, you will need to invest at least around 3.200 USD. Too expensive for a student like me, and my lab hasn't allocated that much money to buy softwares. Anyway, for the solution, I tried to search for freewares. To my surprise, I found a lot of docking freewares. Among them are ArgusLab 4.0, Autodock, and DOCK.

At first, I tried ArgusLab 4.0, which was fairly easy to use. Installation was easy (it runs on windows XP), it has a nice GUI, and accompanied with an easy step by step tutorial, I was able to complete the docking process in no time (you can check out their website: However, the software isn't yet recognized in the scientific community. A quick search using the key word “Arguslab” in google scholar only resulted in 66 hits. There is even a comparison of Arguslab with GOLD, which shows that there are many things that needs to be improved in the current Arguslab version. Thus, I moved on to the other options.

Autodock was indeed tempting, however, the procedures to obtain a copy was somewhat “difficult”. You will need to send a fax or mail with the head of your institutes' signature on the license agreement. I am studying in Japan, and it takes time to complete procedures like that. That's why I moved on to DOCK 6.1. It wasn't hard to obtain, since you only have to fill in an on line request (with permission of your supervising Professor) to get it. I got the software after 2-3 days of my request. Lighting quick response.

So, I got the software, now what? Install it right? Well, it turned out not as easy as I suspected. To install DOCK, you will need a UNIX like environment. On windows, you can use a program called CYGWIN, which generates a UNIX like environment in windows. However, I opted to install openSuse linux (which is UNIX based) on my computer (I had a double OS booting). I compared the docking performance on both (cygwin and linux), and found that it was much more stable using the later (linux). The installation itself was sort of a trial and error process. You will need to install the compilers and other programs that are required to use DOCK (check the tab “dock related links” on their website). For a tutorial on installation of DOCK, you can see the DOCK manual (you will need to improvise though). I will make my own step by step installation soon.

Finished installing, you will have all the programs that you need in the /bin/ directory of your installation directory. However, you will need another program called “dms” (to compute molecular surface), which is not included in the DOCK package (you can get dms from: I'll give details on installing it later.

There are 4 steps in docking with DOCK, which I will mention in my next post.