Why to use a scaling factor for both vdw and electrostatics in QM/MM calculations

[binju...@gmail.com]

I saw most of QM/MM input use the scaling factor for both vdw ( VDW_SCALE14 0.5)  and  electrostatics  ( EI_SCALE14 0.8333333).

It is not clear to me why we need to do this, since we did not us the scaling factor in the classical MD, and the forcefield works well in the classical MD.

Now, we invoke the same forcefield in QM/MM calculation, but we need to scale the vdw and  electrostatics ?  Thanks!

[Dawid das]

Hello,

Whether you scale the 1-4 interctions or not depends on the force field and system (lipid, protein, nucleic

acid). For instance, I use CHARMM22 force field and I do not scale the 1-4 interactions (VDW_SCALE14 and EI_SCALE14 set to 1.0).

You need to read how your force field was optimized and what scaling factor (if any) is suggested.

Also I do not really see reason to do rescaling for QM/MM and not for classical simulations. Plausibly

this rescaling is just related to FF type.

Best wishes,

Dawid

[binju...@gmail.com]

Dear Dawid, Thanks for you answer, I am using the standard Amber forcefield.

best

binju

在 2018年6月6日星期三 UTC+8下午5:15:03，Dawid das写道：

Hello,

Whether you scale the 1-4 interctions or not depends on the force field and system (lipid, protein, nucleic

acid). For instance, I use CHARMM22 force field and I do not scale the 1-4 interactions (VDW_SCALE14 and EI_SCALE14 set to 1.0).

You need to read how your force field was optimized and what scaling factor (if any) is suggested.

[binju...@gmail.com]

Now, I got some clue from Amber manual, By default, vdW 1-4 interactions are scaled down by a factor of 2.0, and the electrostatic 1-4 terms by a factor of 1.2.

在 2018年6月6日星期三 UTC+8下午5:22:23，binju...@gmail.com写道：

[Dawid das]

Yep, so this constitutes to what you wrote in your first message (1/2 =0.5, and 1/1.2 = 0.833333)