Project files provided as supporting information to the manuscript "Information-theoretical measures identify accurate low-resolution representations of protein configurational space"

The dataset contains the following compressed folder:

-Notebooks.zip:

This folder contains:
  -python_script:
        -RESREL.py: script performing the clusterization and computing the relevance resolution curves
        -random_curves.py: script generating the random value and computing the corresponding RES-REV curves_s
        -Cluster_distance_matrix.py: script returning the distance among clusters for a given partition.
  -python_notebook:
        -Exploratory_analysis.ipynb:  Analysis performed on the 12-protein_dataset
        -DMAPS_ANTI.ipynb: Diffusion Map for the Antibody
        -DMAPS_COV_1ake.ipynb: Diffusion Map + Inter-Intra state decomposition of covariance for 1ake

Packages required for the usage of these python scripts/notebooks:
  -numpy
  -pandas
  -matplotlib
  -seaborn
  -multiprocessing
  -scipy

========
RAW DATA
========

The raw data produced and employed in this study are available on a Google Drive folder at the following address:

https://drive.google.com/drive/folders/1PasAUCgpR5-gdzUVEdyusgZIayQN0Le9

In this folder, together with the compressed Notebooks.zip folder, one can fin the compressed folder Data.zip, within which the following data are present:

-12-protein_dataset:
    -md.mdp: the .mdp file used in the MD simulations
    -PROTEIN_PDB_CODE:
        -Hk_{sel}.npy & Hs_{sel}.npy: the Rel & Res curves, sel=[all, CA, CB]
        -RMSD_{sel}.npy: the RMSD matrix, sel=[all, CA, CB]
        -npt.gro:protein+water+ions structure @TEO the equilibration (NVT+NPT)
        -MSR_df.csv: a dataset containing the following columns
                  'area' : area behind the Relevance-Resolution curve;
                  'selection': the atomic selection (['all', 'CA', 'CB']) used to compute the RMSD matrix used for the clusterization (and consequently the Relevance-Resolution curves)
                  'method': the linkage measure used in the clustering procedure, an integer in [0,6];
                  'method_name': the linkage measure used in the clustering procedure, a string in ['average','ward','complete','single','centroid','median','weighted'];
                  'rmsd_mean': the mean value of the rmsd vector along the trajectory computed wrt the first frame;
                  'rmsd_var': the variance of the rmsd vector along the trajectory computed wrt the first frame;
                  'rgy_mean': the mean value of the radius of gyration  along the trajectory;
                  'rgy_var': the variance of the radius of gyration  along the trajectory;
                  'rmsf_mean': the mean value of the rmsf;
                  'rmsf_var': the variance of the rmsf;
                  'RMSD_M_mean': the mean value of the RMSD matrix.
                  'RMSD_M_var': the variance of the RMSD matrix.
  -Random:
    -curves.npy= 100K Relevance-Resolution Random curves for M=40001
    -curves_s.npy= 100K Relevance-Resolution Random curves for M=15000
  -validation_dataset:
    -antibody:
        -Hk_CB.npy & Hs_CB.npy: the Rel & Res curves
        -RMSD_CB.npy: the RMSD matrix
        -DIFF_{M}.npy: the eigenvalue/vector of the 10-D diffusion space
        -Label_{method}.npy: the label vector for n_clusters
    -1ake:
        -Hk_{sel}.npy & Hs_{sel}.npy: the Rel & Res curves
        -RMSD_{sel}.npy: the RMSD matrix
        -DIFF_{M}.npy: the eigenvalue/vector of the 10-D diffusion space
        -Label_{method}.npy: the label vector for n_clusters
        -intra_{m}.npy: the intra-cluster covariance matrix
        -inter_cov_{m}.npy: the inter-cluster correlation matrix

NOTE
=====

The matrices of the cluster distances for adenylate kinase and antibody have been computed through the script Cluster_distance_matrix.py.

These matrices have not been included in the dataset because of their large size; the raw data are however available upon request.