mubind.models.Mubind

class mubind.models.Mubind(datatype, **kwargs)[source]

Implements the MUBIND model.

Args:

datatype (String): Type of the experimental data. “selex” and “pbm” are supported.

Keyword Args:

n_rounds (int): Necessary for SELEX data: Number of rounds to be predicted. init_random (bool): Use a random initialization for all parameters. Default: True padding_const (double): Value for padding DNA-seqs. Default: 0.25 use_dinuc (bool): Use dinucleotide contributions (not fully implemented for all kind of models). Default: False enr_series (bool): Whether the data should be handled as enrichment series. Default: True n_batches (int): Number of batches that will occur in the data. Default: 1 ignore_kernel (list[bool]): Whether a kernel should be ignored. Default: None. kernels (List[int]): Size of the binding modes (0 indicates non-specific binding). Default: [0, 15] n_kernels (int). Number of filters to be used (including non-specific binding, as a constant).

Default: 2 (ns-binding, and one filter)

init_random (bool): Use a random initialization for all parameters. Default: True n_proteins (int): Number of proteins in the dataset. Either n_proteins or n_batches may be used. Default: 1

bm_generator (torch.nn.Module): PyTorch module which has a weight matrix as output. add_intercept (bool): Whether an intercept is used in addition to the predicted binding modes. Default: True

__init__(datatype, **kwargs)[source]: Initialize internal Module state, shared by both nn.Module and ScriptModule.

Methods

`__init__`(datatype, **kwargs)	Initialize internal Module state, shared by both nn.Module and ScriptModule.
`add_module`(name, module)	Add a child module to the current module.
`apply`(fn)	Apply `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])	Return an iterator over module buffers.
`children`()	Return an iterator over immediate children modules.
`compile`(args, *kwargs)	Compile this Module's forward using `torch.compile()`.
`corr_etas_libsizes`(train)
`cpu`()	Move all model parameters and buffers to the CPU.
`cuda`([device])	Move all model parameters and buffers to the GPU.
`dirichlet_regularization`()
`double`()	Casts all floating point parameters and buffers to `double` datatype.
`eval`()	Set the module in evaluation mode.
`exp_barrier`([exp_max])
`extra_repr`()	Set the extra representation of the module.
`float`()	Casts all floating point parameters and buffers to `float` datatype.
`forward`(mono, **kwargs)	Define the computation performed at every call.
`get_buffer`(target)	Return the buffer given by `target` if it exists, otherwise throw an error.
`get_extra_state`()	Return any extra state to include in the module's state_dict.
`get_ignore_kernel`()
`get_kernel_weights`(index, **kwargs)
`get_kernel_width`(index)
`get_log_activities`()
`get_log_etas`()
`get_parameter`(target)	Return the parameter given by `target` if it exists, otherwise throw an error.
`get_submodule`(target)	Return the submodule given by `target` if it exists, otherwise throw an error.
`half`()	Casts all floating point parameters and buffers to `half` datatype.
`ipu`([device])	Move all model parameters and buffers to the IPU.
`load_state_dict`(state_dict[, strict, assign])	Copy parameters and buffers from `state_dict` into this module and its descendants.
`loss_exp_barrier`(exp_max)	We add an exponential negative term, to force weights to be more positive than negative
`loss_kernel_rel`([log])	Return a loss associated to the similarity of weights that are assumed to be similar
`loss_kernel_symmetrical_weights`()	This loss calculates the squared sum of columns per position, and it is useful to detect strong positive/negative biases per position or in the whole object.
`loss_log_dynamic`()
`loss_prob_act`()
`make_model`(train, n_kernels, criterion[, ...])
`modify_kernel`([index, shift, expand_left, ...])
`modules`()	Return an iterator over all modules in the network.
`named_buffers`([prefix, recurse, ...])	Return an iterator over module buffers, yielding both the name of the buffer as well as the buffer itself.
`named_children`()	Return 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])	Return 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, ...])	Return an iterator over module parameters, yielding both the name of the parameter as well as the parameter itself.
`optimize_iterative`(train[, max_w, n_epochs, ...])
`optimize_modified_kernel`(train[, shift, ...])
`optimize_simple`(dataloader, optimiser[, ...])
`optimize_width_and_length`(train, ...[, ...])	A variation of the main optimization routine that attempts expanding the filter of the model at position i, and refines the weights and loss in order to find a better convergence.
`parameters`([recurse])	Return an iterator over module parameters.
`print_weights`()
`register_backward_hook`(hook)	Register a backward hook on the module.
`register_buffer`(name, tensor[, persistent])	Add a buffer to the module.
`register_forward_hook`(hook, *[, prepend, ...])	Register a forward hook on the module.
`register_forward_pre_hook`(hook, *[, ...])	Register a forward pre-hook on the module.
`register_full_backward_hook`(hook[, prepend])	Register a backward hook on the module.
`register_full_backward_pre_hook`(hook[, prepend])	Register a backward pre-hook on the module.
`register_load_state_dict_post_hook`(hook)	Register 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)	Add a parameter to the module.
`register_state_dict_pre_hook`(hook)	Register a pre-hook for the `state_dict()` method.
`requires_grad_`([requires_grad])	Change if autograd should record operations on parameters in this module.
`set_extra_state`(state)	Set extra state contained in the loaded state_dict.
`set_ignore_kernel`(ignore_kernel)
`set_kernel_weights`(weight, index)
`set_seed`(seed, index[, max_value, min_value])
`share_memory`()	See `torch.Tensor.share_memory_()`.
`state_dict`(*args[, destination, prefix, ...])	Return a dictionary containing references to the whole state of the module.
`to`(args, *kwargs)	Move and/or cast the parameters and buffers.
`to_empty`(*, device[, recurse])	Move the parameters and buffers to the specified device without copying storage.
`train`([mode])	Set the module in training mode.
`type`(dst_type)	Casts all parameters and buffers to `dst_type`.
`update_grad`(index, value)
`update_grad_activities`(index, value)
`update_grad_etas`(value)
`weight_distances_min_k`([min_k, exp_delta])
`xpu`([device])	Move all model parameters and buffers to the XPU.
`zero_grad`([set_to_none])	Reset gradients of all model parameters.

Attributes

`T_destination`
`call_super_init`
`dump_patches`
`training`