AMBER for Dummies

(Acknowledgment: Thanks to Ms Lee Hui Jun for giving me a crash course on AMBER)

Setup

Remember to append these lines to your .bashrc:

export PATH=$PATH:/usr/local/amber/exe

export LD_LIBRARY_PATH=$LD_LIBRARY_PATH:/opt/intel/mkl/lib/32

export AMBERHOME=/usr/local/amber

Note: You have to change the path according to where AMBER is installed

In the following step through example, I use p53.pdb as an example

Step Through

• Open Cygwin window, type ‘startx’
• Type ‘xhost +<ipaddress>’, where <ipaddress> is the address of the server you are going to run AMBER. Example: xhost +viper.bii-sg.org
• At your computer, check your own ipaddress (Start->Run->type ‘cmd’; At command prompt, type ‘ipconfig’; note down the IP Address)
• Log into the server where you run AMBER
• After you log in, type ‘export DISPLAY=<IP Address>:0’, where <IP Address> is the address that you noted down in step 3)
• Type ‘xleap’ and you should see xleap window
• At xleap window:
• source leaprc.ff94
• p=loadpdb p53.pdb [Initialize object p to p53]
• solvatebox p TIP3PBOX 8 [add water to p where the box size is 8; you may change to 10 or 12]
• edit p [You will see your protein in a solvatebox. Check if your protein is entirely in the box.]
• check p [Check net charge of p. In this case, p53 is –ve charge]
• addions p Na+ 0 [If –ve charge, add Na+, if +ve charge, add Cl-; the last digit 0 indicates we want net charge of 0]
• saveamberparm p p53.top p53.crd [save protein in solvatebox into .top and .crd]
• quit
• You are supposed to check the residue ID of the last residue of the protein and the total number of residues in p. Download amber_script.sh.
• Amber_script.sh will do the following:
• Generate p53_box.pdb and check for last residue of protein and total number of residues in p
• It then generates .in and .s files for minimization, heating, equilibration and production run and also a runmd file for submitting the jobs through LSF.
• Consider your .top and .crd files are in p53 directory
• Usage: amber_script.sh <directory> <filename>
• Example: amber_script.sh p53 p53
• Files generated:
• p53_min1.in, p53_min1.s, Initial energy minimization with solute frozen
• p53_min2.in, p53_min2.s, Initial energy minimization with solvent frozen
• p53_min3.in, p53_min3.s, Initial energy minimization on whole system
• p53_heat1.in,p53_heat1.s, Heat from 0 to 100 (10ps)
• p53_heat2.in, p53_heat2.s, Heat from 100 to 200 (10ps)
• p53_heat3.in, p53_heat3.s, Heat from 200 to 300 (10ps)
• p53_md2.in, p53_md2.s, Equilibration (100ps)
• p53_md<X>3.in, p53_md<X>.s, Production run (2ns each) [p53_md3 – p53_md7 are generated for the purpose of running 5 x 2ns]
• p53_runmd, just type ‘p53_runmd’ and it will run right from p53_min1.s till p53_md7.s in sequence
• Alternatively, if LSF is not working, you can do prun (Go to your .s file and copy and paste from there; Copy from sander command till end of double quotes)
• Example: prun –n 16 /usr/local/amber/exe/sander -O -i 1H24BE6_min1.in -o 1H24BE6_min1.out -p 1H24BE6.top -c 1H24BE6.crd -r 1H24BE6_min1.rst -ref 1H24BE6.crd &
• Note: In this case, you will have to check if each job has finished running before starting the next one.
• You may want to check the progress by checking the .out files. For example, ‘tail -40 1H24BE6_min1.out’

RMSF (rms.in)

Edit rms.in by changing <file>_mdx.crd to whichever crd files that you want to analyze.

(Content of rms.in)

trajin <file>_md3.crd

trajin <file>_md4.crd

rms first out cyca_mol.rms @N,C,CA time 1.0

To check rmsf against simulation time, type ‘ptraj <file>.top < rms.in’

Output: <file>.rms

Extract frames from trajectories (getframe.in)

Edit getframe.in by changing <file>_mdx.crd to whichever crd files that you want to extract frames.

The trajin line includes ‘trajin <crd file> <start frame> <end frame> <frame interval>.

In this case, 15 frames will be extracted from md3 and md4 files each. (Total of 30 frames.)

The frames extracted are output as pdb file without water molecules.

(Content of getframe.in)

trajin <file>_md3.crd 1 15000 1000

trajin <file>_md4.crd 1 15000 1000

trajout <file>_mdx.pdb pdb nowat

strip :WAT

go

To get the frames, type ‘ptraj <file>.top < getframe.in’.

Output: <file>.pdb.xxx where xxx is the frame number.

MM_PBSA for Dummies

To calculate binding free energy, you have to calculate enthalpy and entropy. But before that, you have to extract the frames from the MD simulations:

Step Through

• Download mm_pbsa.zip
• Edit the file pbsa.sh, change the following definitions:
• WKGDIR, which is the working directory
• PBSA_DIR, which is the directory which stores the files that you downloaded
• PDB_FILE, which is the pdb filename (without file extension)
• PREFIX, which is the prefix of the frames that you extracted (prefix for naming of output files)
• COMPT, RECPT, LIGPT (The top and crd files of complex, receptor and ligand in vacuum.)
• TRAJ (The md production run trajectories crd files)
• NSTART, NSTOP, NFREQ (1^st frame and last frame no. of each trajectory file and interval of frames which has to be used.)
• Type ‘pbsa.sh’. This will create GF, ENTHALPY, ENTROPY_GF and ENTROPY and decomp directories. All the scripts needed are in the corresponding directories. A run_pbsa.sh file will be created.
• Type ‘run_pbsa.sh’. This will do calculation on enthalpy and entropy.

Energy decomposition

• To do energy decomposition, you must have the .top and .crd files for the complex, receptor and ligand and name them as <file>_vac.

For example, 1H24B.pdb is the receptor, 1H24E is the ligand and 1H24BE is the complex, then you should use xleap or tleap to generate 1H24B_vac.top and 1H24B_vac.crd, 1H24E_vac.top and 1H24E_vac.crd, 1H24BE_vac.top and 1H24BE_vac.crd.

To do this, you can do the following:

· tleap

· source leaprc.ff94

· p=loadpdb 1H24B.pdb

· saveamberparm p 1H24B_vac.top 1H24B_vac.crd

· q=loadpdb 1H24B.pdb

· saveamberparm q 1H24B_vac.top 1H24B_vac.crd

· r=loadpdb 1H24BE.pdb

· saveamberparm r 1H24BE_vac.top 1H24BE_vac.crd

· quit

Then go to decomp directory and type ‘bsub –n 1 –q long /usr/local/amber/exe/mm_pbsa.pl decomp.in’.