| version 1.8 | version 1.9 |
|---|
| |
| {\it 54506 Vand\oe uvre--l\`es--Nancy cedex, France} | {\it 54506 Vand\oe uvre--l\`es--Nancy cedex, France} |
| \end{quote} | \end{quote} |
| | |
| | Further additions have been contributed by: Jordi Cohen |
| | |
| | |
| \subsubsection{Introduction and theoretical background} | \subsubsection{Introduction and theoretical background} |
| |
| {5.} | {5.} |
| {This is a radius-shifting coefficient of $\lambda$ that is used | {This is a radius-shifting coefficient of $\lambda$ that is used |
| to construct the modified vdW interactions during alchemical FEP. Providing a positive value for {\tt fepVdWShiftCoeff} ensures that the vdW potential is finite everywhere for small values of $\lambda$, which significantly improves the accuracy and convergence of FEP calculations, and also prevents overlapping particles from making the simulation unstable. During FEP, the inter-atomic distances used in the Lennard-Jones potential are shifted | to construct the modified vdW interactions during alchemical FEP. Providing a positive value for {\tt fepVdWShiftCoeff} ensures that the vdW potential is finite everywhere for small values of $\lambda$, which significantly improves the accuracy and convergence of FEP calculations, and also prevents overlapping particles from making the simulation unstable. During FEP, the inter-atomic distances used in the Lennard-Jones potential are shifted |
| according to: \\ | according to (assuming $\lambda = 0$ means no interaction): \\ |
| $r^2 \rightarrow r^2 + {\rm fepVdWShiftCoeff} \times (1. - \lambda)$ | $r^2 \rightarrow r^2 + {\tt fepVdWShiftCoeff} \times (1 - \lambda)$ |
| } | } |
| | |
| \item | \item |
| \NAMDCONFWDEF{fepVdwScaleExp}{\FEP\ Lennard-Jones parameter scaling exponent} | \NAMDCONFWDEF {fepElecLambdaStart}{\FEP\ lambda start-point for electrostatics} |
| {decimal} | {positive decimal} |
| {0.} | {0.} |
| {When constructing the modified vdW interactions during alchemical FEP, the Lennard-Jones parameters are scaled according to:\\ | {In order to avoid the FEP ``end-point catastrophe", it is often important to make sure that a growing particle does not have an unbounded potential right at the moment that it is created (in case that it appears on top of another particle). One way to deal with this for electrostatic interactions, is to allow a bounded scaled vdW potential (using a positive {\tt fepVdWShiftCoeff}) to first repel all overlapping particles at low values of $\lambda$. As $\lambda$ increases, once the particles are repelled, it is now safe to turn on FEP electrostatics. {\tt fepElecLambdaStart} is the value of the scaling factor $\lambda$ at which electrostatic interactions are turned on and start ramping up linearly; below this value, electrostatics are turned off for all FEP particles. Note: what is really being modified is the $\lambda$ factor that directly multiplies the interactions (\emph{i.e.}, 0 = fully off, 1 = fully on), rather than the $\lambda$ provided by the user, which may sometimes correspond to (1 - $\lambda$).} |
| $A \rightarrow A \times \lambda^{2 \times {\rm fepVdwScaleExp}}$ \\ | |
| $B \rightarrow B \times \lambda^{\rm fepVdwScaleExp}$ | |
| } | \item |
| | \NAMDCONFWDEF {fepVdwLambdaEnd}{\FEP\ lambda end-point for van der Waals} |
| | {positive decimal} |
| | {1.} |
| | {If one is using the {\tt fepElecLambdaStart} option above, one may wish to further decouple the scaling of van der Waals and electrostatic interactions. {\tt fepVdwLambdaEnd} sets the value of $\lambda$ above which all vdW interactions will be fully on. Note: what is really being modified is the $\lambda$ factor that directly multiplies the interactions (\emph{i.e.}, 0 = fully off, 1 = fully on), rather than the $\lambda$ provided by the user, which may sometimes correspond to (1 - $\lambda$).} |
| | |
| %\item | %\item |
| %\NAMDCONFWDEF{fepElecLambdaDelay}{\FEP\ lambda ``delay" for electrostatics} | %\NAMDCONFWDEF {fepAnnihilate}{Make the FEP moiety completely vanish.} |
| %{positive decimal} | %{{\tt on} or {\tt off}} |
| %{0.} | %{{\tt on}} |
| %{In order to avoid the FEP ``end-point catastrophe", it is often important to make sure that a growing particle does not have an unbounded potential right when it is created (in case that it appears on top of another particle). One way to deal with this for electrostatic interactions, is to allow a bounded scaled vdW potential (using a positive fepVdWShiftCoeff) to first repel all overlapping particles at low values of $\lambda$. As $\lambda$ increases, once the particles are repelled, it is now safe to turn on FEP electrostatics. fepElecLambdaDelay is the value of $\lambda$ at which electrostatic interactions are turned on and start ramping up linearly.} | %{{\tt fepAnnihilate} provides the choice between either annihilating the FEP moieties or simply decoupling them from their environment. If {\tt fepAnnihilate} is set to {\tt on}, as $\lambda$ goes to zero, all the interaction energies and forces involving the FEP moiety are scaled. If {\tt fepAnnihilate} is set to {\tt off}, only the interactions between the FEP moiety and its environment are scaled, and all internal energies and interactions remain at their full strength. In certain cases, turning off {\tt fepAnnihilate} directly measures, say, the solvation energy, and would obviate the need to run a separate FEP simulation of the FEP moiety in vacuum to establish its self-energy.} |
| | |
| | |
| | |
| | |
| \end{itemize} | \end{itemize} |