Sugar transporter mechanism

SP superfamily transport cycle

The Project consisted of three main parts:

Machine Learning using Python/Tensorflow

a. ML_svm contains the machine learning script used for training and visualizing the linear SVM training.
b. neural_network contains the architecture and training regime of the network along with the trained weights.
c. inputs contains the inputs necessary for the ML_svm training, whereas the training set for the neural network would be all of the available sugar transporter PDBs as of January 2022.

Structure Prediction using RosettaMP

a. Inputs contains the HYBRIDIZE.zip file which in turn contains the RosettaMP inputs for all conformational states, each of which have a dedicated folder in this OSF project directory. For each point, relevant input files and crucial results are shared.

Enhanced sampling on GLUT5

a. awh_xvgs contains the necessary outputs to analyse convergence etc.
b. mdp files contain the input settings for GROMACS
c. topology and gro files contains the topology, parameters and structures for running the simulations.