SDA FAQ

SDA
Simulation of Diffusional Association
(version 6)

Frequently asked questions

I do not have UHBD, can I still run SDA ?

No-force simulations (or simulations with only hydrophobic forces) need no electrostatic potential grids. These simulations will do Brownian dynamics of solute molecules that are not interacting electrostatically. In these simulations, formation of different types of encounter complexes can be monitored and the shape of the solutes at atomic level accuracy will also be taken into account.
To simulate proteins interacting electrostatically, you need electrostatic potential grids for your proteins in UHBD format. These can be calculated with UHBD or with other programs, for example, APBS. Important to note is that electrostatic potential grids calculated with APBS and written in UHBD format are in ASCII format and in the units kT/e, whereas potentials written by UHBD are in units of kcal/mol/e. To make computed potentials with APBS written in UHBD format compatible with SDA, you may use the program auxi/apbs2uhbd, with an appropriate scaling parameter for converting kT to kcal/mole (0.6 at 300 K).

I do not have NACCESS, can I still run SDA ?

SDA can use the NACCESS subroutine solva() instead of a slow routine with the same name in the distribution package of SDA. When tested on pdb file 1lzt, 1000 atoms, SDA's solva.f routine appears to be 6 times slower than NACCESS and has a maximal deviation of 1.2 Å² and average deviation 0.2 Å² for atomic accessibilities calculated with the probe radius of 1.4 Å. For a larger number of atoms, the accuracy of accessibility calculations does not change, but SDA's solva.f calculation time increases proportionally to the square of the number of atoms. For a dimer of pyruvate kinase, 1a49, ca 16000 atoms, NACCESS calculations take 5.5 sec and SDA's solva.f calculations take 550 sec, on a 2 GHz CPU. To switch to the NACCESS routine, copy subroutines solva() and sortag() of the NACCESS file accall.f (lines 699-1076) to the file solva.f and recompile SDA. Pay attention to matching the values of maxa in maxdim.inc and maxs in this new solva.f - .

I do not have HARLEM, can I still run SDA to calculate electron transfer rates ?

You might use a simpler description of electron transfer than is implemented in SDA (via many possible pathways). In some cases, good estimates can be obtained by simply giving electron donor and acceptor as reaction atom pairs for SDA. >

Should I always use ECM to run SDA ?

Instead of effective charges you can use other charges, keeping in mind that electrostatic interactions calculated using them may be less accurate, than interactions calculated using effective charges.
To check the accuracy of interaction energies calculated with SDA you can use the program sda-ener for a set of representative complexes and compare these energies with separately calculated energies (for example, with UHBD).

What are the 2 variants of effective charges ?

Variant _R of effective charges is used in SDA6. To obtain the electrostatic potential of a solute in solvent using these charges, it is necessary to divide by the solvent relative dielectric constant. For example, in the case of a charge in a nonionic solvent, the potential is: effective_charge_R/distance/dielectric_constant_of_solvent.
Variant _E charges are input to SDA 4.23 and they are smaller than variant _R charges by a factor equal to the solvent dielectric constant. For the same example, the potential is: effective_charge_E/distance.
One more example: effective charge values for a unit charge at the center of a spherical cavity of low dielectric constant in water with dielectric constant 78.5 are: _R variant: 1.0, _E variant: 1/78.5~0.012739.

How do I change solvent ionic strength conditions in SDA calculations ?