Training samples
- 500 ns all atom dataset: https://github.com/tamarabidone/multiscale-model-alphaIIbBeta3
- Link to all-atom paper: https://doi.org/10.1016/j.bpj.2022.12.032
Notebooks for all-atom to CG mapping and diffusion map analysis
- all_atom_cg_mapping.ipynb: Notebook to map all-atom 3D configurations to 300 bead CG configurations.
-
cg_300_integrated_training_integrin_dmap_heterokernel_gridsearch.ipynb: Notebook to identify
$\alpha$ and bandwidth$\varepsilon$ in density-adaptive diffusion maps using grid search. - cg_300_integrated_training_integrin_dmap.ipynb: Notebook to perform density-adaptive diffusion map analysis. Also contains procedure used to remove translational and rotational variance between frames in all four metastable states using Kabsch algorithm to prepare dataset for cWGAN training.
Unified-low dimensional embedding of the gross structure relations between and within metastable states of integrin: evecs_evals_integrin_aa_cg_300.pkl