Create `TrainingInstance`s from raw amino acid sequences.

Computes cumulative quantiles from curriculum weights.

Returns a tensor of weights that correspond to the current curriculum, as parametrized by base and slope. Args: base: Value of the base

Imports configuration dictionary from disk

Convert protein dict into TFRecord.

Takes triplets of dihedral angles (phi, psi, omega) and returns 3D points ready for use in reconstruction of coordinates. Bond lengths and an

Computes the dRMSD of two tensors of vectors. Vectors are assumed to be in the third dimension. Op is done element-wise over batch. Arg

Looks up words embeddings for id tensor. Args: input_ids: int32 Tensor of shape [batch_size, seq_length] containing word ids.

Performs various post-processing on a word embedding tensor. Args: input_tensor: float Tensor of shape [batch_size, seq_length, emb

Make a set of empty features for no-template evalurations

Function used to add C-Beta to glycine resides input: 3 coords (a,b,c), (L)ength, (A)ngle, and (D)ihedral output: 4th coord

Constructs a `BertConfig` from a Python dictionary of parameters.

Gathers the vectors at the specific positions over a minibatch.

Gathers the vectors at the specific positions over a minibatch.

Maps a string to a Python function, e.g., "relu" => `tf.nn.relu`. Args: activation_string: String name of the activation function. Ret

Gets a collection of `InputExample`s for the dev set.

Calculates is chimeric loss Similar implementation to get_global_is_chunk_permuted_output.

Calculates chunk permutation loss Similar implementation to get_global_is_chimeric_output.

Calculates global perturbation loss.

Gets the list of labels for this data set.

Get loss and log probs for the masked LM.

Get loss and log probs for the masked LM.

Get loss and log probs for the next sentence prediction.

Gets a collection of `InputExample`s for prediction.

Gets a collection of `InputExample`s for the train set.

Returns handles to ops (values) of variables by name that can be evaluated by sess.run() Args: vars_: list of strings containin

Returns a boolean mask corresponding to the chosen id filter from a list of ids

Runs layer normalization followed by dropout.

convert string of letters to list of ints

Create an AlphaFold model runner name -- The name of the model to get the parameters from. Options: model_[1-5]

Create a feature dictionary for input to AlphaFold runner - The model runner being invoked. Returned from `make_model_runner` sequence -

Constructs a masking matrix to zero out pairwise distances due to missing residues or padding. This function needs to be called for each in

Returns `model_fn` closure for TPUEstimator.

Returns `model_fn` closure for TPUEstimator.

Returns `model_fn` closure for TPUEstimator.

Returns `model_fn` closure for TPUEstimator.

Package AlphFold's output into an easy-to-use dictionary prediction_result - output from running AlphaFold on an input dictionary process

Read in a PDB file from a path

Takes points from dihedral_to_point and sequentially converts them into the coordinates of a 3D structure. Reconstruction is done in paralle

Reads and parses a protein TF Record. Primary sequences are mapped onto 20-dimensional one-hot vectors. Evolutionary sequences are

Read a Mathematica protein record from file and convert into dict.

Computes the weighted mean of angles. Accepts option to compute use complex exponentials or real numbers. Complex number-based version is gi

Ingest a decoy protein, pass it to AlphaFold as a template, and extract the parsed output target_seq -- the sequence to be predicted deco

Takes torsion_and_curvature_to_rotation_translation and sequentially converts them into the coordinates of a 3D structure. This version

Performs tensor multiplication, broadcasting over the outer dimension of A. Args: Aijk = A 3D tensor with shape [i,j,k]

Serializes this instance to a Python dictionary.

Takes torsion and curvature angles (psi $equiv$ x and theta $equiv$ y) and returns a rotation and translation matrix (based on Niemi et al. 2011). Ea

Multi-headed, multi-layer Transformer from "Attention is All You Need". This is almost an exact implementation of the original Transformer encode

Takes a vector of weights and returns a weighting matrix in which the ith weight is in the ith upper diagonal of the matrix. All other entri

Create TF example files from `TrainingInstance`s. Adapted from https://github.com/google- research/bert/blob/master/create_pretraining_data.py

Run the cell with the declared zoneouts.

Constructs a InputExample. Args: guid: Unique id for the example. text_a: string. The untokenized text of the first seque

Constructs a AdamWeightDecayOptimizer.

Constructs BertConfig. Args: vocab_size: Vocabulary size of `inputs_ids` in `BertModel`. hidden_size: Size of the encoder

Constructor for BertModel. Args: config: `BertConfig` instance. is_training: bool. true for training model, false for eva

Loads configuration from disk and calls concrete method to assign values to local attributes

Create a cell with hidden state and memory zoneout. If this class is used to wrap a Dropout cell, then it will override the output Dropout b

Sets up type of instance object and invokes TF graph creation function.

Return the match method once, then stop

Returns the current global step.

Compute and return diagnostic measurements like weight norms and curriculum quantiles.

Evaluates loss(es) and returns dicts with the relevant loss(es).