1146-1-3bqfr-PosePredictionProtocol.txt

Name

GrAMBina

Software

GrAMBina, AMBER, Gromacs, Chimera, RDKIT

System Preparation Parameters

Assumed pH 6.0
GAFF charges

System Preparation Method

Homology models (Using Modeller) of BACE receptors containing macrocycles were generated to remove all missing ressidues. The structures with the ligand bound to the receptors were minimized using Gromacs 2018 and the latest AMBER forcefield. System Preparation Method: These receptors were used in an ensemble docking protocol. System Preparation Method: Ligands were generated using RDKIT (max 10-conformers). System Preparation Method: In order to generate binding poses for ligands, the macrocycle was broken and the connecting atoms were replaced with "Glue" atoms. System Preparation Method: xs-optimized radii of atoms (based on the Gromacs/AMBER minimized poses) in the receptor, ligand and GrAMBina package (a modified smina version) were used for pose prediction. System Preparation Method: A strong gaussian potential was added so that these Glue atoms are next to each other in the final pose with a length of 1.51 A apart. System Preparation Method: Best poses were visually selected.

Pose Prediction Parameters

Iterated Local Search global optimizer search method
Exhaustiveness=24 #exhaustiveness of global search (default=8)
GrAMBina scoring funtion (empirical + knowledge-based function)
Num_modes=1 #max number of poses to generate
Energy_range=5 #energy difference (kcal/mol) between the best and worst binding mode

Pose Prediction Method

After docking best poses were visually selected.

Answer 1

Yes

Answer 2

No

1146-4-ix868-FreeEnergyProtocol_FE.txt

Name

g_mmpbsa

Software

Gromacs, g_mmpbsa, acpype

Parameter

AMBER force field
GAFF Amber14
1ns
Neutral
10x frames

Method

The system was heated to 310 K (τ_t = 0.2) during a 500-­ps constant-­volume simulation with 2-­fs time step using the modified Berendsen thermostat (V-­rescale) using velocity rescaling. The pressure was equilibrated to 1 atm during a 1,000-­ps constant-­pressure simulation with a 2-­fs time step using the Parrinello–Rahman parameters for pressure coupling. In both simulations, all heavy atoms were position-­restrained with the force constant of 1,000 kJ/(mol nm 2 ). For the molecular dynamics run, the temperature and pressure were maintained at 310 K and 1 atm using the Berendsen thermostat (V-­
rescale) and Parrinello–Rahman pressure coupling method. The short-­range non-­bonded interactions were computed for the atom pairs within the cutoff of 1 nm, and the long-­range electrostatic interactions were calculated using particle-­mesh-­Ewald summation method with fourth-­order cubic interpolation and 0.12 nm grid spacing. The parallel Linear Constraint Solver (LINCS) method was used to constrain bonds. The g_mmpbsa package with the molecular mechanics Poisson–Boltzmann surface area (MM-­PBSA) software from the gromacs and apbs packages was used to calculate the binding energy. The output energies were normalized to between -7.27 and 13.5 kcal/mol, a weakest (5000 nM) and strongest (0.1 nM) binder.

Answer 1

Yes

Answer 2

No