DrugEncoder.DNN

Click here to view source code.

class DNN(in_dim: int, ft_dim: int, hid_dim: int = 512, num_layers: int = 2, dropout: float = 0.3)

DNN can be used to encode ECFPs of drugs.

PARAMETERS:

• in_dim (int) - Dimension of preprocessed ECFPs.

• ft_dim (int) - Dimension of encoded ECFPs.

• hid_dim (int, optional) - Dimension of hidden layers. (default: 512)

• num_layers (int, optional) - Number of torch.nn.Linear layers. (default: 2)

• dropout (float, optional) - Dropout rate of DNN. (default: 0.3)

SHAPES:

• input: Preprocessed ECFPs [batch_size, in_dim]

• output: Encoded ECFPs [batch_size, ft_dim]

forward(f: torch.Tensor)

• f (torch.Tensor) - The input of DNN.