docs: document plugin mechanisms and deepmd-gnn (#4345)

<!-- This is an auto-generated comment: release notes by coderabbit.ai --> ## Summary by CodeRabbit - **New Features**	- Introduced plugin mechanisms for external models in the DeePMD-kit. - Added comprehensive instructions for creating models using PyTorch and a framework-independent backend. - **Documentation** - Enhanced organization of the third-party documentation for better clarity and navigability. <!-- end of auto-generated comment: release notes by coderabbit.ai --> Signed-off-by: Jinzhe Zeng <jinzhe.zeng@rutgers.edu>

Author: njzjz

README.md

@@ -74,6 +74,7 @@ See [our latest paper](https://doi.org/10.1063/5.0155600) for details of all fea
 
 - Multiple backends supported. Add PyTorch and JAX backends.
 - The DPA-2 model.
+- Plugin mechanisms for external models.
 
 ## Install and use DeePMD-kit
 

doc/development/create-a-model-pt.md

@@ -180,6 +180,7 @@ The arguments here should be consistent with the class arguments of your new com
 ## Package new codes
 
 You may package new codes into a new Python package if you don't want to contribute it to the main DeePMD-kit repository.
+A good example is [DeePMD-GNN](https://github.com/njzjz/deepmd-gnn).
 It's crucial to add your new component to `project.entry-points."deepmd.pt"` in `pyproject.toml`:
 
 ```toml

doc/third-party/out-of-deepmd-kit.md

@@ -2,11 +2,21 @@
 
 The codes of the following interfaces are not a part of the DeePMD-kit package and maintained by other repositories. We list these interfaces here for user convenience.
 
-## OpenMM plugin for DeePMD-kit
+## Plugins
+
+### External GNN models (MACE/NequIP)
+
+[DeePMD-GNN](https://github.com/njzjz/deepmd-gnn) is DeePMD-kit plugin for various graph neural network (GNN) models.
+It has interfaced with [MACE](https://github.com/ACEsuit/mace) (PyTorch version) and [NequIP](https://github.com/mir-group/nequip) (PyTorch version).
+It is also the first example to the DeePMD-kit [plugin mechanism](../development/create-a-model-pt.md#package-new-codes).
+
+## C/C++ interface used by other packages
+
+### OpenMM plugin for DeePMD-kit
 
 An [OpenMM](https://github.com/openmm/openmm) plugin is provided from [JingHuangLab/openmm_deepmd_plugin](https://github.com/JingHuangLab/openmm_deepmd_plugin), written by the [Huang Lab](http://www.compbiophysics.org/) at Westlake University.
 
-## Amber interface to DeePMD-kit
+### Amber interface to DeePMD-kit
 
 Starting from [AmberTools24](https://ambermd.org/), `sander` includes an interface to the DeePMD-kit, which implements the [Deep Potential Range Corrected (DPRc) correction](../model/dprc.md).
 The DPRc model and the interface were developed by the [York Lab](https://theory.rutgers.edu/) from Rutgers University.
@@ -17,18 +27,20 @@ More details are available in
 - [DP-Amber](https://github.com/njzjz/dpamber/), a tiny tool to convert Amber trajectory to DPRc training data;
 - [The original DPRc paper](https://doi.org/10.1021/acs.jctc.1c00201).
 
-## CP2K interface to DeePMD-kit
+### CP2K interface to DeePMD-kit
 
 [CP2K](https://github.com/cp2k/cp2k/) v2024.2 adds an interface to the DeePMD-kit for molecular dynamics. Read the [CP2K manual](https://manual.cp2k.org/trunk/methods/machine_learning/deepmd.html#deepmd-kit) for details.
 
-## DP-GEN
+### ABACUS
 
-[DP-GEN](https://github.com/deepmodeling/dpgen) provides a workflow to generate accurate DP models by calling DeePMD-kit's command line interface (CLI) in the local or remote server. Details can be found in [this paper](https://doi.org/10.1016/j.cpc.2020.107206).
+[ABACUS](https://github.com/deepmodeling/abacus-develop/) can run molecular dynamics with a DP model. User is required to [build ABACUS with DeePMD-kit](https://abacus.deepmodeling.com/en/latest/advanced/install.html#build-with-deepmd-kit).
 
-## MLatom
+## Command line interface used by other packages
 
-[Mlatom](http://mlatom.com/) provides an interface to the DeePMD-kit within MLatom's workflow by calling DeePMD-kit's CLI. Details can be found in [this paper](https://doi.org/10.1007/s41061-021-00339-5).
+### DP-GEN
 
-## ABACUS
+[DP-GEN](https://github.com/deepmodeling/dpgen) provides a workflow to generate accurate DP models by calling DeePMD-kit's command line interface (CLI) in the local or remote server. Details can be found in [this paper](https://doi.org/10.1016/j.cpc.2020.107206).
 
-[ABACUS](https://github.com/deepmodeling/abacus-develop/) can run molecular dynamics with a DP model. User is required to [build ABACUS with DeePMD-kit](https://abacus.deepmodeling.com/en/latest/advanced/install.html#build-with-deepmd-kit).
+### MLatom
+
+[Mlatom](http://mlatom.com/) provides an interface to the DeePMD-kit within MLatom's workflow by calling DeePMD-kit's CLI. Details can be found in [this paper](https://doi.org/10.1007/s41061-021-00339-5).