i think you are making some mistake with preparing your complex(the protein and the best pose for the ligand )

i advice you to try charm GUI website for preparing your complex automatically http://www.charmm-gui.org/

go to input generator and then for solution builder and then follow the steps its easy.

@Anas Alobaidi

Can't thank you enough

You helped me a lot

Hi Heba Shawky ,

how did you define those position restraints. I'd recommend doing another minimization without constraints, so the hbonds etc. can actually form between protein and ligand. This is because the docking structures do usually not contain hydrogens and when building the system, the actual contacts may not be well formed.

Regarding the error with the cell, I assume you are using NAMD and in this case for initial NPT simulations just set "margin" in your script to prevent it from crashing. I usually set it "margin 2", but values up to 5 may be reasonable if the initial density is far from desired. Notice that this makes the simulations slower, therefore only set it when required, this is, during the first NPT run where the cell shrinks most.

Away from all that, please also consider that "one" simulation will not tell you much about the stability of the ligand binding, It is totally expected that ligands bind and unbind in an equilibrium distribution. Actually in theory, from that distribution of bound vs unbound frames you can directly derive the binding free energy based on Boltzmann's law. I'd recommend to do at least multiple simulations considering the additional unconstrained minimization.

Are you planning to utilize MM/GBSA afterwards?

Norman Geist

Hello sir, thank you for replying

let me in brief explain what i'm doing

I'm testing protein carriers for prodigiosin as a drug , four protein molecules, one is modeled and three crystal structures

I have docked them using autodock vina, generated my complexes using UCSF Chimera and went for my first simulation ever.

I followed the NAMD-VDM tutorial just to learn..this tutorial includes the minimization of the whole system , then a position constraint MD , fixing the whole complex atoms except hydrogens and ultimately a free long production run for 20 ns. As for my work, I'm honsetly cluless what exactly should I do to assess my complex stability. I have read a lot of papers but each one has its own protocol arrangements..So, if you can guide me what steps should I go for i'll be very grateful.

one more thing, as you mentioned NPT equilibriation; I have generated an NPT config file as it's not provided in tutorial files, could you please have a look and tell me if this setup is o.k? i've also attached the minimization and position constraint config files to look at them if you wouldn't mind

Thank a million in advance

Regarding the minimize.namd:

- For some reason it contains a langevinpiston section (NPT) ^^

- Instead of PMEGridSizeXYZ simply set "PMEGridSpacing 1", then you dont need to change it all the time.

- Try increasing steps for minimization.

- The minimization is already without constraints, so everything fine here

sim_fixprot.namd:

- It is already NPT, that's why it crashes without "margin"

- Same here regarding PMEGridSize

- Increase the run time to several nanoseconds

-Dont set "seed" manually

-What is actually constrained here?

npt.conf:

- Remove cellBasisVecor123 since you already use the restart information "extendedsystem"

-Dont use flexibleCell and useconstantarea unless you have a membrane spanning in XY direction

nvt.conf

- This is far fro what you want, it actually uses a spherical box ^^

- Just adapt the nvt.conf and remove langevinpiston

TIP: Don't use any "wrap" command of NAMD, it's much much much much easier to wrap in VMD afterwards:

pbc wrap -all -compound fragment -center com -centersel "protein"

Norman Geist

You wouldn't know how much I'm grateful for such detailed guidance..Thanks a lot..

just one more question and please excuse me being a neoby :)

you said earlier concerning position constraint MD " Dont set "seed" manually " ..i didn't..as I earlier mentioned I'm using the tutorial provided configuration files and tailor them for my work, but it seems like I drift away in some setup...

should I remove the seeding or how can I set it up?

sorry again for asking too much:)

Sry for the late reply. What I was refering to is the "seed" keyword wich is set in some of your input files. It's generally not necessary or advisable to set this random seed value manually, especially if you are not satisfied with the timeline the simulation took from the initial velocity distribution, this is, the ligand leaving the binding pocket. While this does not mean that the same seed will lead to exactly reproduced trajecotries, timelines will diverge much slower if you use the same seed everytime. NAMD will choose a own random number at startup which makes your results more reasoable in case you run multiple tries to enhance statistics.

Norman Geist no need ever to be sorry..it's me who can't thank you enough for sharing your great expertise with me..

surely i'll consider each point you refereed to..

many thanx again

wishing you all the best

Hello. Did you use CHARMM-GUI server in order to get the files for the simulation successfully ran? I'm stuck preparing files in CHARMM-GUI, as the PDB reader always shows errors for my complex.pdb file.