"gr Predictor" is a deep-learning model rapidly estimating the water-oxygen distribution function around protein obtained by the 3D-RISM theory from the protein 3D structure. The computation is completed within a few minutes when using a single CPU and approximately one minute when using a single GPU.

Python 3.6–, Anaconda, openbabel, dask, tesorflow

“gr Predictor” is available under the GNU General Public License.

If you use “gr Predictor”, please cite:

gr Predictor: a Deep-Learning Model for Predicting the Hydration Structures around Proteins 
Kosuke Kawama, Yusaku Fukushima, Mitsunori Ikeguchi, Masateru Ohta, and Takashi Yoshidome
Journal of Chemical Information and Modeling, accepted for publication (2022).

If you have any questions, please contact Takashi Yoshidome at [email protected].

• Import modules

  import guv_prediction as gpred

• Prediction of g(r)

  spliter = gpred.Prediction( pdb_dir, dx_dir, pred_area_center = np.array([x, y, z]), pred_area_range = w )

  • pdb_dir : Path to the PDB file

  • dx_dir : Path to the output file (dx format)

  • pred_area_center : Central coordinate of the prediction region in the protein ( x, y, and z [angstrom] )

  • pred_area_range : Width of the prediction region ( w [angstrom] )