antipasti.model package

Submodules

antipasti.model.model module

This module contains the model class.

Authors:

Kevin Michalewicz <k.michalewicz22@imperial.ac.uk>

class antipasti.model.model.ANTIPASTI(*args: Any, **kwargs: Any)[source]

Bases: Module

Predicting the binding affinity of an antibody from its normal mode correlation map.

Parameters:
  • n_filters (int) – Number of filters in the convolutional layer.

  • filter_size (int) – Size of filters in the convolutional layer.

  • pooling_size (int) – Size of the max pooling operation.

  • input_shape (int) – Shape of the normal mode correlation maps.

  • l1_lambda (float) – Weight of L1 regularisation.

  • mode (str) – To use the full model, provide full. Otherwise, ANTIPASTI corresponds to a linear map.

forward(x)[source]

Model’s forward pass.

Returns:

  • output (torch.Tensor) – Predicted binding affinity.

  • inter_filter (torch.Tensor) – Filters before the fully-connected layer.

l1_regularization_loss()[source]

Module contents

This subpackage contains the model class.