1. The number of resultant simulations of the peptide in surface conformation is not three. Why is that?
Three simulations are performed starting with a TM peptide orientation, and three with a surface orientation.
However, during the simulation, the peptide is allowed to move (translate and rotate) in the membrane, and
simulations starting with TM orientation may end in surface orientations if the peptide is amphipathic, too short
to span the membrane or simply too polar. The opposite situation, i.e., converting from surface to TM orientation is
rare because of the high energy barrier associated with the transfer of the polar peptide termini from the aqueous
phase into the membrane. In addition, simulations that produce conformations of positive free energy on average are
discarded, so that the total number of resultant simulations may be smaller than the total number of initial
simulations (i.e., six).
(back to top)
2. What is the estimated time of a simulation of my peptide?
The run time depends on the length of your peptide. Approximated run times for peptides of 10-50 amino acids are presented in the following table:
(back to top)
|Number of amino acids
||Estimated run time (minutes)|
3. Is there a correlation between the results of MCPep and Molecular Dynamics (MD) simulations?
In our previous studies we correlated the data obtained via the MC model used in MCPep to the available data, either experimental or derived from MD simulations. For example:
- Kessel et al. (2003) simulated M2δ, a model peptide corresponding to the second TM segment of the acetylcholine δ-subunit.
The conformations obtained from the MC simulations were very similar to the structures determined by a nuclear magnetic resonance study
(Oppella et al., 1999),
with average root-mean-square deviation of less than 3 Å. Moreover, the calculated value of the tilt angle of the TM helical peptide of 14°
correlated well with the experimental value of 12° (Opella et al., 1999).
- Shental-Bechor et al. (2007) simulated the antimicrobial peptide Magainin2 within a membrane containing 30% of charged (and 70% zwitterionic) lipids.
The calculated membrane-association free energy was -11.16 ± 1.0 kT. The binding energy estimated from the
MD simulations was -13.8 kT
The binding energy estimated experimentally was -11.6 kT (Wieprecht et al.,1999),
in agreement with both MC and MD simulations.
(back to top)
- Gordon-Grossman et al. (2009) simulated the antimicrobial peptide melittin in membranes containing various lipid compositions.
In neutral membrane the computed value -15.1 ± 0.6 kT was slightly lower than the experimental value of about -11.8 to -13.6 kT
(Ladokhin and White, 2001).
We also compared our results to MD simulations of melittin within a membrane containing 10% anionic lipids
The free energy value of melittin-membrane association in the MD simulation was -21.7 kT in excellent agreement with the -21.1 ± 0.6 kT value of our MC
4. I provided a structure of my peptide in Protein Data Bank format. Why is it changed?
Your pdb file is changed to remove unnecessary additions (e.g., heteroatoms), so that only the coordinates of the
standard amino acids remain. When the pdb entry contains multiple model-structures, as in nuclear magnetic resonance
studies, only the first model is used. (back to top)