Inputs needed for preparation:

• density maps (one or more binned MRC or .rec files)
• coordinate files (corresponding coordinates file. *.mod or *.coords or *.pts)

Output you will get:

Root directory

— Autopick:

— Autopick_[NAME]:

— data: folder storing data used to train the neural network

— test_x

— test_y

— train_x

— train_y

— metrics.png: image storing model metric

Explanation for each of the required fields in the GUI:

• Input Folder: Input folder for network training.
• Continue from model: If you already have a model, use this to continue training that model.
• Output folder name: Folder to store your output model.
• epoch #: Number of complete passes through the entire training dataset.
• GPU ID: ID of the GPUs you plan on using.
• Subtomo #: The subtomogram number to be extracted from each of the tomograms.
• Subtomo box size: Box size of the subtomograms.
• Training y label size: The y label size used in generating training dataset target Y.

ADVANCED

• learning rate: The learning rate for the training network
• batch size: Batch size for the training network
• steps per epoch: Steps per epoch for the training network