Dynamic PDB: A New Dataset and a SE(3) Model Extension by Integrating Dynamic Behaviors and Physical Properties in Protein Structures

Ce Liu^1*  Jun Wang^1*  Zhiqiang Cai^1*  Yingxu Wang^1,3  Huizhen Kuang²  Kaihui Cheng²  Liwei Zhang¹ 

Qingkun Su¹  Yining Tang²  Fenglei Cao¹  Limei Han²  Siyu Zhu^2†  Yuan Qi^2† 

¹Shanghai Academy of Artificial Intelligence for Science  ²Fudan University 
³Mohamed bin Zayed University of Artificial Intelligence

Overview

Dynamic PDB is a large-scale dataset that enhances existing prestigious static 3D protein structural databases, such as the Protein Data Bank (PDB), by integrating dynamic data and additional physical properties. It contains approximately 12.6k filtered proteins, each subjected to all-atom molecular dynamics (MD) simulations to capture conformational changes.

Compared with previously existing protein MD datasets, dynamic PDB provides three key advancements:

1. Extended simulation durations: Up to 1 microsecond per protein, facilitating a more comprehensive understanding of significant conformational changes.
2. Finer-grained sampling intervals: 1 picosecond intervals, allowing for the capture of more detailed allosteric pathways.
3. Enriched array of physical properties: Captured during the MD process, including atomic velocities and forces, potential/kinetic energies, and the temperature of the simulation environment, etc.

What dynamic PDB contains?

The attributes contained in dynamic PDB are listed as follows:

File Name	Attribute	Data Type	Unit
{protein_id}_T.pkl	Trajectory coordinates	float array	Å
{protein_id}_V.pkl	Atomic velocities	float array	Å/ps
{protein_id}_F.pkl	Atomic forces	float array	kcal/mol·Å
{protein_id}_npt_sim.dat	Potential energy Kinetic energy Total energy Temperature Box volume System density	float float float float float float	kJ/mole kJ/mole kJ/mole K nm³ g/mL

In addition, the following data are stored during the MD simulation:

File Name	Description
{protein_id}_minimized.pdb	PDB structure after minimization
{protein_id}_nvt_equi.dat	Information in NVT equilibration
{protein_id}_npt_equi.dat	Information in NPT equilibration
{protein_id}_T.dcd	DCD format for trajectory coordinates
{protein_id}_state_npt1000000.0.xml	Status file for MD prolongation

Data Availability Notice

Thank you for your interest and support in our dataset. Due to the immense size of the full simulation data and storage limitations, we have decided to provide the 100ns simulation data for all proteins for online download, with detailed usage instructions provided below (Download Dataset). For researchers who require the 1µs simulation data for specific proteins, please fill out an agreement to the dynamicPDB Terms of Use, using your institutional email addresses, and send it to us at [email protected], along with the PDB IDs. We will provide the data individually. The full list of PDB IDs can be downloaded here.

Download Dataset

You can easily get dynamic PDB dataset from our ModelScope repo.

1. Make sure you have Git LFS installed:

sudo apt-get install git-lfs
# Initialize Git LFS
git lfs install

2. Navigate to your DATA_ROOT and clone the source:

GIT_LFS_SKIP_SMUDGE=1 git clone https://www.modelscope.cn/datasets/fudan-generative-vision/dynamicPDB.git dynamicPDB_raw

GIT_LFS_SKIP_SMUDGE=1 configures Git to clone the pointers for all LFS files.

3. Download data with a specific protein_id, for example 1a62_A:

cd dynamicPDB_raw
git lfs pull --include="{protein_id}/*"

4. Merge the split-volume compression into one file and then unzip the .tar.gz file:

cat {protein_id}/{protein_id}.tar.gz.part* > {protein_id}/{protein_id}.tar.gz
cd ${Your Storage Root}
mkdir dynamicPDB  # ignore if directory exists
tar -xvzf dynamicPDB_raw/{protein_id}/{protein_id}.tar.gz -C dynamicPDB

Finally, the dataset should be organized as follows:

./dynamicPDB/
|-- 1a62_A_npt100000.0_ts0.001
|   |-- 1a62_A_npt_sim_data
|   |   |-- 1a62_A_npt_sim_0.dat
|   |   `-- ...
|   |-- 1a62_A_dcd
|   |   |-- 1a62_A_dcd_0.dcd
|   |   `-- ...
|   |-- 1a62_A_T
|   |   |-- 1a62_A_T_0.pkl
|   |   `-- ...
|   |-- 1a62_A_F
|   |   |-- 1a62_A_F_0.pkl
|   |   `-- ...
|   |-- 1a62_A_V
|   |   |-- 1a62_A_V_0.pkl
|   |   `-- ...
|   |-- 1a62_A.pdb
|   |-- 1a62_A_minimized.pdb
|   |-- 1a62_A_nvt_equi.dat
|   |-- 1a62_A_npt_equi.dat
|   |-- 1a62_A_T.dcd
|   |-- 1a62_A_T.pkl
|   |-- 1a62_A_F.pkl
|   |-- 1a62_A_V.pkl
|   `-- 1a62_A_state_npt100000.0.xml
|-- 1ah7_A_npt100000.0_ts0.001
|   |-- ...
|   `-- ...
`-- ...

For ease of use, we have also provided segmented versions of the data (directories {protein_id}_dcd, {protein_id}_T, {protein_id}_F, and {protein_id}_V), each representing one-tenth of the total simulation duration, sequentially named from 0 to 9 in chronological order. The files {protein_id}_T.dcd, {protein_id}_T.pkl, {protein_id}_F.pkl, {protein_id}_V.pkl are their corresponding combination.

Applications

1. Physics condition: we extend the SE(3) diffusion model to incorporate sequence features and physical properties for the task of trajectory prediction.
2. 4D diffusion: we introduce a 4D diffusion with motion alignment to learn dynamic protein structures.

Acknowledgements

We would like to thank the contributors to the OpenFold, OmegaFold.
If we missed any open-source projects or related articles, we would like to complement the acknowledgement of this specific work immediately.