Starting structures for the cpH-MD simulations were prepared from the available crystal structures on the protein data bank, using the structures 5EVE (Amb a 8), 5EM0 (Art v 4) and 5NZB (Bet v 2) (30 (link)). All residues not corresponding to the actual allergen itself were removed during setup. Topologies and starting coordinates were prepared with the tLEaP module of AmberTools 20 (Case et al.), using the ff99SB force field (31 (link)), along with modification necessary for cpH-MD (32 (link)–34 (link)). Generalized Born (GB) radii of the titratable oxygens in the aspartate and glutamate side chains were reduced to 1.3 Å, as suggested by Swails et al. (34 (link)). Each system was soaked in a truncated octahedral box of TIP3P (35 (link)) water with a minimum wall distance of 10 Å. All systems were equilibrated with an extensive protocol before production (36 (link), 37 (link)).
Starting structures for the GaMD simulations were extracted from the obtained cpH-MD trajectories as follows: For each system, at each simulated pH value, the trajectories were clustered into 5 clusters with the program cpptraj of AmberTools 20 (38 ) using a hierarchical agglomerative approach and average linkage. Each cluster structure was then set up for subsequent GaMD simulations with the program tLEaP (38 ) using the ff14SB force field (39 (link)) and a cubic TIP3P water box, with 10 Å padding.
Starting structures for the GaMD simulations were extracted from the obtained cpH-MD trajectories as follows: For each system, at each simulated pH value, the trajectories were clustered into 5 clusters with the program cpptraj of AmberTools 20 (38 ) using a hierarchical agglomerative approach and average linkage. Each cluster structure was then set up for subsequent GaMD simulations with the program tLEaP (38 ) using the ff14SB force field (39 (link)) and a cubic TIP3P water box, with 10 Å padding.
Free full text: Click here
