From 25135c3b59b5bc7faeec5e4c7fae56af359ca335 Mon Sep 17 00:00:00 2001 From: "Matthew W. Thompson" Date: Mon, 18 Jul 2022 13:43:43 -0500 Subject: [PATCH 1/7] Remove `long_range_dispersion` from vdW section samples --- docs/standards/smirnoff.md | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/docs/standards/smirnoff.md b/docs/standards/smirnoff.md index fd7597e..f568e1d 100644 --- a/docs/standards/smirnoff.md +++ b/docs/standards/smirnoff.md @@ -317,7 +317,7 @@ As an example of a complete SMIRNOFF force field specification, see [a recent fo van der Waals force parameters, which include repulsive forces arising from Pauli exclusion and attractive forces arising from dispersion, are specified via the `` tag with sub-tags for individual `Atom` entries, such as: ```XML - + ... @@ -339,7 +339,7 @@ Support for [other Lennard-Jones mixing schemes](https://en.wikipedia.org/wiki/C Later revisions will add support for additional potential types (e.g., `Buckingham-exp-6`), as well as the ability to support arbitrary algebraic functional forms using a scheme such as ```XML - + From 2cbc2febc65e28fc6533ba596b62ed613a143d96 Mon Sep 17 00:00:00 2001 From: "Matthew W. Thompson" Date: Mon, 18 Jul 2022 14:42:23 -0500 Subject: [PATCH 2/7] Draft OFF-EP 0007 --- docs/enhancement-proposals/off-ep-0007.md | 92 +++++++++++++++++++++++ docs/standards/smirnoff.md | 7 +- 2 files changed, 97 insertions(+), 2 deletions(-) create mode 100644 docs/enhancement-proposals/off-ep-0007.md diff --git a/docs/enhancement-proposals/off-ep-0007.md b/docs/enhancement-proposals/off-ep-0007.md new file mode 100644 index 0000000..34ea7e4 --- /dev/null +++ b/docs/enhancement-proposals/off-ep-0007.md @@ -0,0 +1,92 @@ +# OFF-EP 0007 — Add long-range dispersion attribute in vdW section + +**Status:** Draft + +**Authors:** Matt Thompson + +**Stakeholders:** + +**Acceptance criteria:** + +**Created:** 2022-06-18 + +**Discussion:** [PR #40](https://github.com/openforcefield/standards/pull/40) + +**Implementation:** [``openff-standards``](https://github.com/openforcefield/openff-standards) + +## Abstract + +This change exposes long-range dispersion corrections as a toggleable option in the vdW section of +the SMIRNOFF specification. + +## Motivation and Scope + +In version 0.3 of the `` tag, there is no exposed option for turning on or off long-range +dispersion corrections. Existing implementaions of SMIRNOFF, to our knowldge, all use an isotropic +correction, but in some use cases it may be desired to turn this off or use a different type of +correction. This proposal makes this explicit and also adds a non-default value that enables users +to turn this correction off. It does not attempt to modify other non-bonded settings (cut-off +distance and method, switching function, mixing rule, or long-range electrostatics treatment). + +## Usage and Impact + +Existing SMIRNOFF implementations currently turn on a long-range dispersion correction (i.e. [OpenFF +Toolkit](https://github.com/openforcefield/openff-toolkit/blob/0.10.6/openff/toolkit/typing/engines/smirnoff/parameters.py#L3695)). +Updating the specification would not impact this behavior if using default and recommended settings. +Users wishing to turn this correction off, for whatever reason, would now be able to do so while +still following the SMIRNOFF specifications. + +Most engines implement some form of isotropic long-range dispersion correction. In +[OpenMM](http://docs.openmm.org/latest/api-python/generated/openmm.openmm.NonbondedForce.html#openmm.openmm.NonbondedForce.setUseDispersionCorrection), +this is controlled via a method that accepts a boolean toggle. Similar options are implemented +in [GROMACS](https://manual.gromacs.org/current/user-guide/mdp-options.html#mdp-DispCorr), AMBER +(`nodisper`), [LAMMPS](https://docs.lammps.org/pair_modify.html) and likely other engines. However +unlikely, if in the future other types of corrections than the current isotropic one are developerd, +this specification can be extended because the setting is a string value, not a boolean. + +## Backward compatibility + +This change proposes adding an attribute to a section, so in a strict sense it is not +backwards-comptaible. (A parser that only knows about version 0.3 might not understand a serialized +representation of a version 0.4 section because of this new attribute.) However, the recommended +default value matches what is currently ubiquituous in SMIRNOFF implementations, so an up-converter +from section 0.3 to 0.4 should be straightforward to write and safe to use. It is recommended that +any 0.3 version section is up-converted to a 0.4 section with the only change being an added +`long_range_dispersion` attribute with its default value of `"isotropic"`. + +## Detailed description + +A `long_range_dispersion` attribute is added to the `vdW` section, which is bumped to 0.4. +Supported values are + * `"isotropic"`: Isotropic dispersion corrections described in + [Shirts, 2007](https://pubs.acs.org/doi/10.1021/jp0735987) should be used. + * `"none"`: No dispersion corrections should be used. + +The default value, which is recommended for general use, is `"isotropic"`. + +These options only sensibly apply to periodic systems and only to vdW interactions. For non-periodic +systems, this attribute should be ignored. + +No other sections are updated, therefore this change is not meant to impact electrostatic interactions. + +## Alternatives + +No alternative proposals are under consideration or offered. If this proposal were to be rejected, +implementations would likely continue using long-range dispersion corrections, but would lack +guidance from the SMIRNOFF specification and no other values (such as turning corrections off) would +be unamiguously supported. + +## Discussion + +- [Standards #38](https://github.com/openforcefield/standards/pull/38): A user discovered that `long_range_dispersion` is not an attribute in version 0.3. +- [OFF-EP 00002](https://github.com/openforcefield/standards/pull/22): A naive proposal that attempted to make this change without noticing that it would be breaking. + +## Copyright + +* This was seeded from the +[OFF-EP template](https://github.com/openforcefield/standards/blob/main/docs/enhancement-proposals/off-ep-template.md), +which was is based upon the +[``numpy`` NEP template]( https://github.com/numpy/numpy/blob/master/doc/neps/nep-template.rst) and the +[``conda-forge`` CFEP template.](https://github.com/conda-forge/cfep/blob/master/cfep-00.md)* + +*All OFF-EPs are explicitly [CC0 1.0 Universal](https://creativecommons.org/publicdomain/zero/1.0/).* diff --git a/docs/standards/smirnoff.md b/docs/standards/smirnoff.md index f568e1d..de97ae3 100644 --- a/docs/standards/smirnoff.md +++ b/docs/standards/smirnoff.md @@ -317,7 +317,7 @@ As an example of a complete SMIRNOFF force field specification, see [a recent fo van der Waals force parameters, which include repulsive forces arising from Pauli exclusion and attractive forces arising from dispersion, are specified via the `` tag with sub-tags for individual `Atom` entries, such as: ```XML - + ... @@ -327,7 +327,9 @@ For standard Lennard-Jones 12-6 potentials (specified via `potential="Lennard-Jo The two are related by `r0 = 2^(1/6)*sigma` and conversion is done internally into the `sigma` values used in OpenMM. Attributes in the `` tag specify the scaling terms applied to the energies of 1-2 (`scale12`, default: 0), 1-3 (`scale13`, default: 0), 1-4 (`scale14`, default: 0.5), and 1-5 (`scale15`, default: 1.0) interactions, -as well as the distance at which a switching function is applied (`switch_width`, default: `"1.0*angstrom"`), and the cutoff (`cutoff`, default: `"9.0*angstroms"`). Version 0.3 assumes that the long-range dispersion handling is isotropic. +as well as the distance at which a switching function is applied (`switch_width`, default: `"1.0*angstrom"`), the cutoff (`cutoff`, default: `"9.0*angstroms"`). + +Version 0.4 includes a `long_range_dispersion` attribute (default: `"isotropic"`) that controls if long-range dispersion correction(s) are to be used. Supported values are `"isotropic"` (an isotroping long-range dispersion correction as desribed in [Shirts, 2007](https://pubs.acs.org/doi/10.1021/jp0735987) should be used) or `"none"` (use none). Version 0.3 assumes that the long-range dispersion handling is isotropic. The `potential` attribute (default: `"none"`) specifies the potential energy function to use. Currently, only `potential="Lennard-Jones-12-6"` is supported: @@ -362,6 +364,7 @@ Later revisions will also provide support for special interactions using the `` From db629feb3e3bc508888324369f72fd741b14584e Mon Sep 17 00:00:00 2001 From: "Matthew W. Thompson" Date: Mon, 18 Jul 2022 15:14:04 -0500 Subject: [PATCH 3/7] Update stakeholders --- docs/enhancement-proposals/off-ep-0007.md | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/docs/enhancement-proposals/off-ep-0007.md b/docs/enhancement-proposals/off-ep-0007.md index 34ea7e4..4002e7b 100644 --- a/docs/enhancement-proposals/off-ep-0007.md +++ b/docs/enhancement-proposals/off-ep-0007.md @@ -4,7 +4,7 @@ **Authors:** Matt Thompson -**Stakeholders:** +**Stakeholders:** Jeffrey Wagner, David Mobley, John Chodera **Acceptance criteria:** From 70e4b0dd0d09848c7a389349fbc6bd9db13d49a3 Mon Sep 17 00:00:00 2001 From: "Matthew W. Thompson" Date: Tue, 19 Jul 2022 08:52:02 -0500 Subject: [PATCH 4/7] Edits, reference toolkit #1351 --- docs/enhancement-proposals/off-ep-0007.md | 20 +++++++++++--------- 1 file changed, 11 insertions(+), 9 deletions(-) diff --git a/docs/enhancement-proposals/off-ep-0007.md b/docs/enhancement-proposals/off-ep-0007.md index 4002e7b..9c7b3d5 100644 --- a/docs/enhancement-proposals/off-ep-0007.md +++ b/docs/enhancement-proposals/off-ep-0007.md @@ -23,15 +23,15 @@ the SMIRNOFF specification. In version 0.3 of the `` tag, there is no exposed option for turning on or off long-range dispersion corrections. Existing implementaions of SMIRNOFF, to our knowldge, all use an isotropic -correction, but in some use cases it may be desired to turn this off or use a different type of -correction. This proposal makes this explicit and also adds a non-default value that enables users +correction, but in some current or future use cases it may be desired to turn this off or use a different type of +correction. This proposal makes this an explicit attribute and also adds a non-default value that enables users to turn this correction off. It does not attempt to modify other non-bonded settings (cut-off distance and method, switching function, mixing rule, or long-range electrostatics treatment). ## Usage and Impact Existing SMIRNOFF implementations currently turn on a long-range dispersion correction (i.e. [OpenFF -Toolkit](https://github.com/openforcefield/openff-toolkit/blob/0.10.6/openff/toolkit/typing/engines/smirnoff/parameters.py#L3695)). +Toolkit](https://github.com/openforcefield/openff-toolkit/blob/0.10.6/openff/toolkit/typing/engines/smirnoff/parameters.py#L3695)) following guidance from [Standards #38](https://github.com/openforcefield/standards/pull/38) recently and a lack of guidance previously. Updating the specification would not impact this behavior if using default and recommended settings. Users wishing to turn this correction off, for whatever reason, would now be able to do so while still following the SMIRNOFF specifications. @@ -40,15 +40,16 @@ Most engines implement some form of isotropic long-range dispersion correction. [OpenMM](http://docs.openmm.org/latest/api-python/generated/openmm.openmm.NonbondedForce.html#openmm.openmm.NonbondedForce.setUseDispersionCorrection), this is controlled via a method that accepts a boolean toggle. Similar options are implemented in [GROMACS](https://manual.gromacs.org/current/user-guide/mdp-options.html#mdp-DispCorr), AMBER -(`nodisper`), [LAMMPS](https://docs.lammps.org/pair_modify.html) and likely other engines. However -unlikely, if in the future other types of corrections than the current isotropic one are developerd, +(`nodisper`), [LAMMPS](https://docs.lammps.org/pair_modify.html), and likely other engines. However +unlikely, if in the future other types of corrections than the current isotropic one are developed, this specification can be extended because the setting is a string value, not a boolean. ## Backward compatibility This change proposes adding an attribute to a section, so in a strict sense it is not -backwards-comptaible. (A parser that only knows about version 0.3 might not understand a serialized -representation of a version 0.4 section because of this new attribute.) However, the recommended +backwards-compatible. (A parser that only knows about version 0.3 might not understand a serialized +representation of a version 0.4 section because of this new attribute, and at very least the +information would be lost in a down-conversion.) However, the recommended default value matches what is currently ubiquituous in SMIRNOFF implementations, so an up-converter from section 0.3 to 0.4 should be straightforward to write and safe to use. It is recommended that any 0.3 version section is up-converted to a 0.4 section with the only change being an added @@ -65,7 +66,7 @@ Supported values are The default value, which is recommended for general use, is `"isotropic"`. These options only sensibly apply to periodic systems and only to vdW interactions. For non-periodic -systems, this attribute should be ignored. +systems, this attribute should be ignored by implementations. No other sections are updated, therefore this change is not meant to impact electrostatic interactions. @@ -74,12 +75,13 @@ No other sections are updated, therefore this change is not meant to impact elec No alternative proposals are under consideration or offered. If this proposal were to be rejected, implementations would likely continue using long-range dispersion corrections, but would lack guidance from the SMIRNOFF specification and no other values (such as turning corrections off) would -be unamiguously supported. +unamiguously be supported. ## Discussion - [Standards #38](https://github.com/openforcefield/standards/pull/38): A user discovered that `long_range_dispersion` is not an attribute in version 0.3. - [OFF-EP 00002](https://github.com/openforcefield/standards/pull/22): A naive proposal that attempted to make this change without noticing that it would be breaking. +- [OpenFF Toolkit #1351](https://github.com/openforcefield/openff-toolkit/pull/1351): A sample implementation with example behavior ## Copyright From 310ade5d4683d59e63bf747d812e865854a6abc8 Mon Sep 17 00:00:00 2001 From: "Matthew W. Thompson" Date: Thu, 21 Jul 2022 13:37:05 -0500 Subject: [PATCH 5/7] Reference OpenMM implementation alongside Shirts paper --- docs/enhancement-proposals/off-ep-0007.md | 3 ++- docs/standards/smirnoff.md | 2 +- 2 files changed, 3 insertions(+), 2 deletions(-) diff --git a/docs/enhancement-proposals/off-ep-0007.md b/docs/enhancement-proposals/off-ep-0007.md index 9c7b3d5..452acdd 100644 --- a/docs/enhancement-proposals/off-ep-0007.md +++ b/docs/enhancement-proposals/off-ep-0007.md @@ -60,7 +60,8 @@ any 0.3 version section is up-converted to a 0.4 section with the only change be A `long_range_dispersion` attribute is added to the `vdW` section, which is bumped to 0.4. Supported values are * `"isotropic"`: Isotropic dispersion corrections described in - [Shirts, 2007](https://pubs.acs.org/doi/10.1021/jp0735987) should be used. + [Shirts, 2007](https://pubs.acs.org/doi/10.1021/jp0735987) and implemented in + [OpenMM](http://docs.openmm.org/latest/api-python/generated/openmm.openmm.NonbondedForce.html#openmm.openmm.NonbondedForce.setUseDispersionCorrection) should be used. * `"none"`: No dispersion corrections should be used. The default value, which is recommended for general use, is `"isotropic"`. diff --git a/docs/standards/smirnoff.md b/docs/standards/smirnoff.md index de97ae3..8e2861a 100644 --- a/docs/standards/smirnoff.md +++ b/docs/standards/smirnoff.md @@ -329,7 +329,7 @@ The two are related by `r0 = 2^(1/6)*sigma` and conversion is done internally in Attributes in the `` tag specify the scaling terms applied to the energies of 1-2 (`scale12`, default: 0), 1-3 (`scale13`, default: 0), 1-4 (`scale14`, default: 0.5), and 1-5 (`scale15`, default: 1.0) interactions, as well as the distance at which a switching function is applied (`switch_width`, default: `"1.0*angstrom"`), the cutoff (`cutoff`, default: `"9.0*angstroms"`). -Version 0.4 includes a `long_range_dispersion` attribute (default: `"isotropic"`) that controls if long-range dispersion correction(s) are to be used. Supported values are `"isotropic"` (an isotroping long-range dispersion correction as desribed in [Shirts, 2007](https://pubs.acs.org/doi/10.1021/jp0735987) should be used) or `"none"` (use none). Version 0.3 assumes that the long-range dispersion handling is isotropic. +Version 0.4 includes a `long_range_dispersion` attribute (default: `"isotropic"`) that controls if long-range dispersion correction(s) are to be used. Supported values are `"isotropic"` (an isotroping long-range dispersion correction as desribed in [Shirts, 2007](https://pubs.acs.org/doi/10.1021/jp0735987) and implemented in [OpenMM](http://docs.openmm.org/latest/api-python/generated/openmm.openmm.NonbondedForce.html#openmm.openmm.NonbondedForce.setUseDispersionCorrection) should be used) or `"none"` (use none). Version 0.3 assumes that the long-range dispersion handling is isotropic. The `potential` attribute (default: `"none"`) specifies the potential energy function to use. Currently, only `potential="Lennard-Jones-12-6"` is supported: From 1fa1a3ecb4ab3bb1b45f6f7e013112a3f48b1345 Mon Sep 17 00:00:00 2001 From: "Matthew W. Thompson" Date: Thu, 21 Jul 2022 13:41:09 -0500 Subject: [PATCH 6/7] Add notes about interaction with LJPME --- docs/enhancement-proposals/off-ep-0007.md | 4 ++-- docs/standards/smirnoff.md | 2 +- 2 files changed, 3 insertions(+), 3 deletions(-) diff --git a/docs/enhancement-proposals/off-ep-0007.md b/docs/enhancement-proposals/off-ep-0007.md index 452acdd..268cc36 100644 --- a/docs/enhancement-proposals/off-ep-0007.md +++ b/docs/enhancement-proposals/off-ep-0007.md @@ -66,8 +66,8 @@ Supported values are The default value, which is recommended for general use, is `"isotropic"`. -These options only sensibly apply to periodic systems and only to vdW interactions. For non-periodic -systems, this attribute should be ignored by implementations. +These options only sensibly apply to periodic systems and only to vdW interactions with a cut-off. For non-periodic +systems or those using full long-range interactions (i.e. LJPME), this attribute should be ignored by implementations. No other sections are updated, therefore this change is not meant to impact electrostatic interactions. diff --git a/docs/standards/smirnoff.md b/docs/standards/smirnoff.md index 8e2861a..6a05802 100644 --- a/docs/standards/smirnoff.md +++ b/docs/standards/smirnoff.md @@ -329,7 +329,7 @@ The two are related by `r0 = 2^(1/6)*sigma` and conversion is done internally in Attributes in the `` tag specify the scaling terms applied to the energies of 1-2 (`scale12`, default: 0), 1-3 (`scale13`, default: 0), 1-4 (`scale14`, default: 0.5), and 1-5 (`scale15`, default: 1.0) interactions, as well as the distance at which a switching function is applied (`switch_width`, default: `"1.0*angstrom"`), the cutoff (`cutoff`, default: `"9.0*angstroms"`). -Version 0.4 includes a `long_range_dispersion` attribute (default: `"isotropic"`) that controls if long-range dispersion correction(s) are to be used. Supported values are `"isotropic"` (an isotroping long-range dispersion correction as desribed in [Shirts, 2007](https://pubs.acs.org/doi/10.1021/jp0735987) and implemented in [OpenMM](http://docs.openmm.org/latest/api-python/generated/openmm.openmm.NonbondedForce.html#openmm.openmm.NonbondedForce.setUseDispersionCorrection) should be used) or `"none"` (use none). Version 0.3 assumes that the long-range dispersion handling is isotropic. +Version 0.4 includes a `long_range_dispersion` attribute (default: `"isotropic"`) that controls if long-range dispersion correction(s) are to be used. Supported values are `"isotropic"` (an isotroping long-range dispersion correction as desribed in [Shirts, 2007](https://pubs.acs.org/doi/10.1021/jp0735987) and implemented in [OpenMM](http://docs.openmm.org/latest/api-python/generated/openmm.openmm.NonbondedForce.html#openmm.openmm.NonbondedForce.setUseDispersionCorrection) should be used) or `"none"` (use none). Version 0.3 assumes that the long-range dispersion handling is isotropic. Note that this correction only applies to periodic systems and those that use cut-off vdW interactions. For example, if LJPME is used, `long_range_dispersion` should be ignored. The `potential` attribute (default: `"none"`) specifies the potential energy function to use. Currently, only `potential="Lennard-Jones-12-6"` is supported: From 2ec761ec6639ade571bdd445e0599033971d9aab Mon Sep 17 00:00:00 2001 From: "Matthew W. Thompson" Date: Thu, 21 Jul 2022 13:52:25 -0500 Subject: [PATCH 7/7] Upwrap bulky paragraph, fix typos --- docs/enhancement-proposals/off-ep-0007.md | 8 ++++---- docs/standards/smirnoff.md | 6 +++++- 2 files changed, 9 insertions(+), 5 deletions(-) diff --git a/docs/enhancement-proposals/off-ep-0007.md b/docs/enhancement-proposals/off-ep-0007.md index 268cc36..20873f6 100644 --- a/docs/enhancement-proposals/off-ep-0007.md +++ b/docs/enhancement-proposals/off-ep-0007.md @@ -16,13 +16,13 @@ ## Abstract -This change exposes long-range dispersion corrections as a toggleable option in the vdW section of +This change exposes long-range dispersion corrections as a toggle-able option in the vdW section of the SMIRNOFF specification. ## Motivation and Scope In version 0.3 of the `` tag, there is no exposed option for turning on or off long-range -dispersion corrections. Existing implementaions of SMIRNOFF, to our knowldge, all use an isotropic +dispersion corrections. Existing implementations of SMIRNOFF, to our knowledge, all use an isotropic correction, but in some current or future use cases it may be desired to turn this off or use a different type of correction. This proposal makes this an explicit attribute and also adds a non-default value that enables users to turn this correction off. It does not attempt to modify other non-bonded settings (cut-off @@ -50,7 +50,7 @@ This change proposes adding an attribute to a section, so in a strict sense it i backwards-compatible. (A parser that only knows about version 0.3 might not understand a serialized representation of a version 0.4 section because of this new attribute, and at very least the information would be lost in a down-conversion.) However, the recommended -default value matches what is currently ubiquituous in SMIRNOFF implementations, so an up-converter +default value matches what is currently ubiquitous in SMIRNOFF implementations, so an up-converter from section 0.3 to 0.4 should be straightforward to write and safe to use. It is recommended that any 0.3 version section is up-converted to a 0.4 section with the only change being an added `long_range_dispersion` attribute with its default value of `"isotropic"`. @@ -76,7 +76,7 @@ No other sections are updated, therefore this change is not meant to impact elec No alternative proposals are under consideration or offered. If this proposal were to be rejected, implementations would likely continue using long-range dispersion corrections, but would lack guidance from the SMIRNOFF specification and no other values (such as turning corrections off) would -unamiguously be supported. +unambiguously be supported. ## Discussion diff --git a/docs/standards/smirnoff.md b/docs/standards/smirnoff.md index 6a05802..49dd534 100644 --- a/docs/standards/smirnoff.md +++ b/docs/standards/smirnoff.md @@ -329,7 +329,11 @@ The two are related by `r0 = 2^(1/6)*sigma` and conversion is done internally in Attributes in the `` tag specify the scaling terms applied to the energies of 1-2 (`scale12`, default: 0), 1-3 (`scale13`, default: 0), 1-4 (`scale14`, default: 0.5), and 1-5 (`scale15`, default: 1.0) interactions, as well as the distance at which a switching function is applied (`switch_width`, default: `"1.0*angstrom"`), the cutoff (`cutoff`, default: `"9.0*angstroms"`). -Version 0.4 includes a `long_range_dispersion` attribute (default: `"isotropic"`) that controls if long-range dispersion correction(s) are to be used. Supported values are `"isotropic"` (an isotroping long-range dispersion correction as desribed in [Shirts, 2007](https://pubs.acs.org/doi/10.1021/jp0735987) and implemented in [OpenMM](http://docs.openmm.org/latest/api-python/generated/openmm.openmm.NonbondedForce.html#openmm.openmm.NonbondedForce.setUseDispersionCorrection) should be used) or `"none"` (use none). Version 0.3 assumes that the long-range dispersion handling is isotropic. Note that this correction only applies to periodic systems and those that use cut-off vdW interactions. For example, if LJPME is used, `long_range_dispersion` should be ignored. +Version 0.4 includes a `long_range_dispersion` attribute (default: `"isotropic"`) that controls if long-range dispersion correction(s) are to be used. +Supported values are `"isotropic"` (an isotropic long-range dispersion correction as described in [Shirts, 2007](https://pubs.acs.org/doi/10.1021/jp0735987) and implemented in [OpenMM](http://docs.openmm.org/latest/api-python/generated/openmm.openmm.NonbondedForce.html#openmm.openmm.NonbondedForce.setUseDispersionCorrection) should be used) or `"none"` (use none). +Note that this correction only applies to periodic systems and those that use cut-off vdW interactions. +For example, if LJPME is used, `long_range_dispersion` should be ignored. +Version 0.3 assumes that the long-range dispersion handling is isotropic. The `potential` attribute (default: `"none"`) specifies the potential energy function to use. Currently, only `potential="Lennard-Jones-12-6"` is supported: