This repo contains the code and files for the paper titled Predicted and Experimental NMR Chemical Shifts at Variable Temperatures: The Effect of Protein Conformational Dynamics (J. Phys. Chem. Lett. 2024, 15, 8, 2270–2278).
The molecular dynamics trajectory, minimized MD snapshots and predicted chemical shifts we used in the paper can be downloaded at here.
- Anaconda
- Schrödinger Suite (Not required for running the minimal example below, but required for reproducing our MD trajectory and minimized snapshots)
- AFNMR
- AmberTools
- ORCA
git clone git@github.com:zlcrrrr/DHFR_MD_QMMM.git
conda env create -n AmberTools22 --file AmberTools22.yml
The DHFR initial structure (DHFR_init.pdb) and the prepared structure for MD (DHFR_md_input.pdb) are in ./initial_structure. The details of the initial structure preparation, and MD simulation setup are in the paper. We provide the 1μs MD trajectory we used in the paper here (~6.75 GB). For reproducing the trajectory, you need to install Schrödinger Suite.
Note: we performed all the AFNMR and QMMM calculations described in the paper on a HPC cluster using Slurm. The scripts provided in the respository are just examples. You may need to tweak our scripts to fit your computing environment.
Download the MD trajectory and snapshots. Unzip and move the contents of the downloaded folders to the repository, and extract MD snapshots from the trajectory to a folder named snapshots
cd sh
chmod +x ./*.sh
./extract_shifts.sh # this requires Schrodinger Suite. But you can use free softwares like VMD to do the same job as well.
Then submit minimization and AFNMR jobs to Slurm.
cd sh
chmod +x ./*.sh
./run_workflow.sh # this scripts requires Schrodinger Suite
Or you can use the minimized snapshots we provided and skip the steps above
cd sh
chmod +x ./*.sh
sbatch run_afnmr.sh
After the minimization and AFNMR jobs are done, run
./prepare_scripts.sh # prepare QM/MM calculation scripts
and
./submit_all.sh # submit QM/MM calculation jobs to Slurm.
Finally, extract the calculated chemical shifts. Results are in a folder ./sh/results/
./extract_shifts.sh
We provide an example to run a AF-QM/MM calculation using AFNMR and ORCA for the first residue (MET1) in the first snapshot of the MD trajectory (schrodinger_md_0.pdb).
Install Anaconda, AFNMR, and ORCA.
Clone the repository and setup the conda environment
git clone git@github.com:zlcrrrr/DHFR_MD_QMMM.git
conda env create -n AmberTools22 --file AmberTools22.yml
Then run AFNMR and QM/MM calculations.
cd ./DHFR_MD_QMMM/example
chmod +x ./run_example.sh
./run_example.sh
The calculated chemical shifts are in results.txt (as shown below). Columns are residue index, atom name, residue name, chemical shift, delta, eta, xx, yy, zz, respectively.
1 N MET 26.962 13.053 0.282 200.985 218.722 222.408
1 H1 MET 4.906 -14.839 0.075 20.321 21.428 43.133
1 H2 MET 5.160 -13.178 0.141 20.522 22.378 41.218
1 H3 MET 4.913 -11.649 0.044 22.208 22.717 39.936
1 CA MET 57.680 19.816 0.690 112.304 135.191 148.864
1 HA MET 3.954 2.103 0.943 27.143 29.305 31.289
1 CB MET 31.892 25.632 0.181 132.276 168.403 173.044
1 HB2 MET 2.319 6.280 0.359 24.601 32.894 35.147
1 HB3 MET 2.331 -6.124 0.734 25.559 30.056 36.993
1 CG MET 27.314 -19.298 0.903 144.127 161.548 181.784
1 HG2 MET 2.205 -6.771 0.866 24.680 30.541 37.766
1 HG3 MET 2.289 6.861 0.925 24.050 31.168 37.514
1 CE MET 20.212 -25.434 0.515 150.322 163.421 195.022
1 HE1 MET 1.603 -10.060 0.321 24.950 28.184 41.657
1 HE2 MET 2.139 -7.020 0.977 24.122 30.979 38.081
1 HE3 MET 1.286 -9.262 0.233 26.202 28.364 41.176
.
├── AmberTools22.yml
├── LICENSE.md
├── README.md
├── bin
│ └── getshifts-orca-old
├── example
│ ├── run_example.sh
│ └── schrodinger_md_0.pdb
├── figures
│ └── fig.png
├── initial_structures
│ ├── DHFR_init.pdb
│ └── DHFR_md_input.pdb
├── lib
│ ├── SO4.frcmod
│ ├── SO4.lib
│ ├── TMP.frcmod
│ └── TMP.lib
├── sh
│ ├── extract_shifts.sh
│ ├── logs
│ ├── prepare_scripts.sh
│ ├── run_afnmr.sh
│ ├── run_minimization.sh
│ ├── run_workflow.sh
│ └── submit_all.sh
└── trj2pdb.sh
7 directories, 20 files