came.CGGCNet

class came.CGGCNet(g_or_canonical_etypes: DGLHeteroGraph | Sequence[Tuple[str]], in_dim_dict: Dict[str, int], h_dim: int = 32, h_dim_add: int | None = None, out_dim: int = 32, num_hidden_layers: int = 1, norm: str = 'right', use_weight: bool = True, dropout_feat: float = 0.0, dropout: float = 0.0, negative_slope: float = 0.05, batchnorm_ntypes: List[str] | None = None, layernorm_ntypes: List[str] | None = None, out_bias: bool = False, rel_names_out: Sequence[Tuple[str]] | None = None, share_hidden_weights: bool = False, attn_out: bool = True, kwdict_outgat: Dict = {}, share_layernorm: bool = True, residual: bool = False, **kwds)

Cell-Gene-Gene-Cell graph neural network.

Graph Convolutional Network for cell-gene Heterogeneous graph, with edges named as:

  • (‘cell’, ‘express’, ‘gene’): ov_adj

  • (‘gene’, ‘expressed_by’, ‘cell’): ov_adj.T

  • (‘gene’, ‘homolog_with’, ‘gene’): vv_adj + sparse.eye(n_gnodes)

  • (‘cell’, ‘self_loop_cell’, ‘cell’): sparse.eye(n_cells)

Notes

  • gene embeddings are computed from cells;

  • weight sharing across hidden layers is allowed by setting share_hidden_weights as True.

  • attention can be applied on the last layer (self.cell_classifier);

  • the graph for the embedding layer and the hidden layers can be different;

  • allow expression values as static edge weights. (but seems not work…)

Parameters:

  • g_or_canonical_etypes (dgl.DGLGraph or a list of 3-length-tuples) – if provide a list of tuples, each of the tuples should be like (node_type_source, edge_type, node_type_destination).

  • in_dim_dict (Dict[str, int]) – Input dimensions for each node-type

  • h_dim (int) – number of dimensions of the hidden states

  • h_dim_add (Optional[int or Tuple]) – if provided, an extra hidden layer will be add before the classifier

  • out_dim (int) – number of classes (e.g., cell types)

  • num_hidden_layers (int) – number of hidden layers

  • norm (str) – normalization method for message aggregation, should be one of {‘none’, ‘both’, ‘right’, ‘left’} (Default: ‘right’)

  • use_weight (bool) – True if a linear layer is applied after message passing. Default: True

  • dropout_feat (float) – dropout-rate for the input layer

  • dropout (float) – dropout-rate for the hidden layer

  • negative_slope (float) – negative slope for LeakyReLU

  • batchnorm_ntypes (List[str]) – specify the node types to apply BatchNorm (Default: None)

  • layernorm_ntypes (List[str]) – specify the node types to apply LayerNorm

  • out_bias (bool) – whether to use the bias on the output classifier

  • rel_names_out (a list of tuples or strings) – names of the output relations; if not provided, use all the relations of the graph.

  • share_hidden_weights (bool) – whether to share the graph-convolutional weights across hidden layers

  • attn_out (bool) – whether to use attentions on the output layer

  • kwdict_outgat (Dict) – a dict of key-word parameters for the output graph-attention layers

  • share_layernorm (bool) – whether to share the LayerNorm across hidden layers

  • residual (bool) – whether to use the residual connection between the embedding layer and the last hidden layer. This operation may NOT be helpful in transfer-learning scenario. (Default: False)

See also

HiddenRRGCN

__init__(g_or_canonical_etypes: DGLHeteroGraph | Sequence[Tuple[str]], in_dim_dict: Dict[str, int], h_dim: int = 32, h_dim_add: int | None = None, out_dim: int = 32, num_hidden_layers: int = 1, norm: str = 'right', use_weight: bool = True, dropout_feat: float = 0.0, dropout: float = 0.0, negative_slope: float = 0.05, batchnorm_ntypes: List[str] | None = None, layernorm_ntypes: List[str] | None = None, out_bias: bool = False, rel_names_out: Sequence[Tuple[str]] | None = None, share_hidden_weights: bool = False, attn_out: bool = True, kwdict_outgat: Dict = {}, share_layernorm: bool = True, residual: bool = False, **kwds)

Initializes internal Module state, shared by both nn.Module and ScriptModule.

Methods

__init__(g_or_canonical_etypes, in_dim_dict)

Initializes internal Module state, shared by both nn.Module and ScriptModule.

add_module(name, module)

Adds a child module to the current module.

apply(fn)

Applies fn recursively to every submodule (as returned by .children()) as well as self.

bfloat16()

Casts all floating point parameters and buffers to bfloat16 datatype.

buffers([recurse])

Returns an iterator over module buffers.

children()

Returns an iterator over immediate children modules.

cpu()

Moves all model parameters and buffers to the CPU.

cuda([device])

Moves all model parameters and buffers to the GPU.

double()

Casts all floating point parameters and buffers to double datatype.

eval()

Sets the module in evaluation mode.

extra_repr()

Set the extra representation of the module

float()

Casts all floating point parameters and buffers to float datatype.

forward(feat_dict, g, **other_inputs)

Defines the computation performed at every call.

get_attentions(feat_dict, g[, fuse])

output a cell-by-gene attention matrix

get_buffer(target)

Returns the buffer given by target if it exists, otherwise throws an error.

get_classification_loss(out_cell, labels[, ...])

get_extra_state()

Returns any extra state to include in the module's state_dict.

get_hidden_states([feat_dict, g, i_layer, ...])

access the hidden states.

get_out_logits(feat_dict, g, **other_inputs)

get the output logits

get_parameter(target)

Returns the parameter given by target if it exists, otherwise throws an error.

get_sampler(canonical_etypes[, ...])

get_submodule(target)

Returns the submodule given by target if it exists, otherwise throws an error.

half()

Casts all floating point parameters and buffers to half datatype.

ipu([device])

Moves all model parameters and buffers to the IPU.

load_state_dict(state_dict[, strict])

Copies parameters and buffers from state_dict into this module and its descendants.

modules()

Returns an iterator over all modules in the network.

named_buffers([prefix, recurse])

Returns an iterator over module buffers, yielding both the name of the buffer as well as the buffer itself.

named_children()

Returns an iterator over immediate children modules, yielding both the name of the module as well as the module itself.

named_modules([memo, prefix, remove_duplicate])

Returns an iterator over all modules in the network, yielding both the name of the module as well as the module itself.

named_parameters([prefix, recurse])

Returns an iterator over module parameters, yielding both the name of the parameter as well as the parameter itself.

parameters([recurse])

Returns an iterator over module parameters.

register_backward_hook(hook)

Registers a backward hook on the module.

register_buffer(name, tensor[, persistent])

Adds a buffer to the module.

register_forward_hook(hook)

Registers a forward hook on the module.

register_forward_pre_hook(hook)

Registers a forward pre-hook on the module.

register_full_backward_hook(hook)

Registers a backward hook on the module.

register_load_state_dict_post_hook(hook)

Registers a post hook to be run after module's load_state_dict is called.

register_module(name, module)

Alias for add_module().

register_parameter(name, param)

Adds a parameter to the module.

requires_grad_([requires_grad])

Change if autograd should record operations on parameters in this module.

set_extra_state(state)

This function is called from load_state_dict() to handle any extra state found within the state_dict.

share_memory()

See torch.Tensor.share_memory_()

state_dict(*args[, destination, prefix, ...])

Returns a dictionary containing a whole state of the module.

to(*args, **kwargs)

Moves and/or casts the parameters and buffers.

to_empty(*, device)

Moves the parameters and buffers to the specified device without copying storage.

train([mode])

Sets the module in training mode.

type(dst_type)

Casts all parameters and buffers to dst_type.

xpu([device])

Moves all model parameters and buffers to the XPU.

zero_grad([set_to_none])

Sets gradients of all model parameters to zero.

Attributes

T_destination

dump_patches