ToPoRg-18k: dataset of single-chain radii of gyration distribution for 18,450 architecturally diverse and chemically patterned coarse-grained polymers

This distribution provides access to 18,450 configurations of coarse-grained polymers. The data is provided as a serialized object using the `pickle' Python module and in csv format. The data was compiled using Python version 3.8. 

ReferencesThe specific applications and analyses of the data are described in 1.  Jiang, S.; Webb, M.A. "Physics-Guided Neural Networks for Transferable Prediction of Polymer Properties"

DataThere are seven .pickle files that contain serialized Python objects.

pattern_graph_data_*_*_rg_new.pickle: squared radii of gyration distribution from MD simulation. The number indicates the molecular weight range.

rg2_baseline_*_new.pickle: squared radii of gyration distribution from Gaussian chain theoretical prediction.

delta_data_v0314.pickle: torch_geometric training data.

UsageTo access the data in the .pickle file, users can execute the following:

# LOAD SIMULATION DATADATA_DIR = "your/custom/dir/"mw = 40 # or 90, 190 MWs

filename = os.path.join(DATA_DIR, f"pattern_graph_data_{mw}_{mw+20}_rg_new.pickle")with open(filename, "rb") as handle:    graph = pickle.load(handle)    label = pickle.load(handle)    desc  = pickle.load(handle)    meta  = pickle.load(handle)    mode  = pickle.load(handle)    rg2_mean   = pickle.load(handle)    rg2_std    = pickle.load(handle) ** 0.5 # var

# combine asymmetric and symmetric star polymerslabel[label == 'stara'] = 'star'# combine bottlebrush and other comb polymerslabel[label == 'bottlebrush'] = 'comb' 

# LOAD GAUSSIAN CHAIN THEORETICAL DATAwith open(os.path.join(DATA_DIR, f"rg2_baseline_{mw}_new.pickle"), "rb") as handle:    rg2_mean_theo = pickle.load(handle)[:, 0]    rg2_std_theo = pickle.load(handle)[:, 0]

graph: NetworkX graph representations of polymers.

label: Architectural classes of polymers (e.g., linear, cyclic, star, branch, comb, dendrimer).

desc: Topological descriptors (optional).

meta: Identifiers for unique architectures (optional).

mode: Identifiers for unique chemical patterns (optional).

rg2_mean: Mean squared radii of gyration from simulations.

rg2_std: Corresponding standard deviation from simulations.

rg2_mean_theo: Mean squared radii of gyration from theoretical models.

rg2_std_theo: Corresponding standard deviation from theoretical models.

Help, Suggestions, Corrections?If you need help, have suggestions, identify issues, or have corrections, please send your comments to Shengli Jiang at [e-mail]

GitHubAdditional data and code relevant for this study is additionally accessible at https://github.com/webbtheosim/gcgnn