https://doi.org/10.5061/dryad.vq83bk42v

Description of the data and file structure

This document outlines the simulation and computational methods utilized in the paper “Free Energy Analysis of Peptide-Induced Pore Formation in Lipid Membranes by Bridging Atomistic and Coarse-Grained Simulations” to aid with reproducibility of results discussed. More detailed information on the workflow, including running umbrella sampling simulations and post-trajectory analysis, are included in the README file in the home directory of the attached tarball. All simulations were run with Gromacs 2021.5 patched with PLUMED 2.8.

1.     The PLUMED nucleation collective variable (ξ) to create aqueous transmembrane pores

•  Included as the file Nucleation_CV.cpp
  • Utilized for all atomistic (AA) and coarse-grained (CG) umbrella sampling simulations
  • AA = CHARMM36 force field
  • CG = MARTINI 2.2 force field
• The following parameters are set for the calculation of ξ based on Hub et. al in J. Chem. Theory Comput. 2017, 13, 5, 2352–2366:
  • Groups:
  • Mem = all membrane atoms (AA) or beads (CG) to calculate the center of the membrane in the z direction
  • Waters = all water oxygens (AA) or water W beads (CG)
  • POxygens = all DMPC lipid head phosphate oxygens (AA) or PO4 beads (CG)
  • NS = defines the number of slices of the transmembrane cylinder
  • DS = slice thickness --> AA = 0.1 nm, CG = 0.2 nm
  • RCYL = radius of transmembrane cylinder --> AA & CG = 0.8 nm
  • ZETA = coefficient that defines slice occupancy after addition of 1 polar atom --> AA & CG = 0.75
  • XCYL and YCYL = defines lateral center of transmembrane cylinder

2.     Umbrella Sampling Simulations

•  All initial CG system files (gro, ndx, top) are provided for the 4 systems considered in this study: Pure DMPC, 8 AUR, 8 MEL, 8 MAG
• The script 'run_cg_setup.sh' is provided which conducts all equilibration and steered MD steps to prepare initial configurations for umbrella sampling
  • 23 windows are considered from ξ = 0.2 (flat membrane) to ξ = 1.0 (fully nucleated pore)
• The script 'run_parallel_us.sh' runs all 23 windows in parallel utilizing the MPI parallelization capabilities of Gromacs.

3.     Backmapping CG to AA configurations

• All mapping (.map) files required to backmap CG systems to their AA representations are provided in the 'backmapping/Mapping' folder
  • DMPC lipids, each peptide amino acid
• The script 'CG_to_AA.sh' is provided to streamline the backmapping process to generate initial configurations for AA umbrella sampling:
  • NOTE: The files backward.py and initram.sh should be downloaded from the Software category provided and placed in the 'backmapping' folder before running CG_to_AA.sh
  • Steps:
  • Backmap CG to AA with backward.py based on initial guesses of atom positions and enforced geometry considerations (e.g., chirality)
  • 2 energy minimization steps (without then with peptide-peptide and membrane-membrane nonbonded interactions turned on)
  • 4 restrained NVT equilibration steps with increasing timestep (∆t = 0.0002, 0.0005, 0.001, 0.002 ps)
  • 2 restrained NPT equilibration steps (∆t = 0.001, 0.002 ps)
  • 1 unrestrained NPT equilibration step (∆t = 0.002 ps)

4.     Trajectory analysis scripts and data

• The following are provided to aid with the reproducibility of Figures 2-7 in the main text:
  • Figure 2 --> Peptide_Density_Convergence.sh and Peptide_Density_Convergence_analysis.py
  • Centers membrane in box, calculates density of peptide in middle 1 nm of pore, converts density to estimated number of pore-lining peptides.
  • Figure 3
  • All PMF vs. ξ files are provided (freefile.txt)
    • Free energies are resolved with Grossfield WHAM --> http://membrane.urmc.rochester.edu/?page_id=126
  • Figure 4 --> num_dens_heatmap.py
  • Calculates average number densities of groups of interest during AA umbrella sampling
    • 0.1 nm radial and 0.1 nm z bins starting from lateral pore center (defined by XCYL and YCYL above)
  • Following groups considered:
    • Lipid head phosphate
    • Water oxygen atoms
    • Peptides
    • Counterions
  • Figure 5a --> pep_helix_angle.py
  • Creates index files defining N-terminal and C-terminal ends of each peptide
  • Calculates average tilt angle of each peptide relative to membrane surface per umbrella sampling window
  • Figure 6a --> SCD.sh
  • Calculates average deuterium order parameters (SCD) of each i-1 to i+1 lipid tail carbon based on gmx order
  • Figure 6c --> SCD_radial.py
  • Similar to SCD.sh, but only calculates the deuterium order parameter of tails with their head phosphate within a 2 nm radius of the pore center (defined by XCYL and YCYL above).
  • Figure 7a
  • All raw salt bridge .dat files are provided, which are the raw data (Figure S25) that were used in the calculation of the 'Normalized Salt Bridge' metric.
    • Salt bridges were calculated with the 'salt bridges' plugin in vmd setting a cutoff distance of 4 Angstroms (0.4 nm) 

Version changes

Oct-2024: Standardized all instances of 'gmx' to 'gmx_mpi' for calling Gromacs for simulation and analysis. Removed some extraneous files (e.g., centered trajectory .xtc) that are generated with provided scripts to reduce .tar download size.