Given a quaternion representing rotation, get the quaternion representing its inverse. Args: quaternion: Quaternions as tensor o

Convert rotations given as quaternions to axis/angle. Args: quaternions: quaternions with real part first, as tensor of

Generate random quaternions representing rotations, i.e. versors with nonnegative real part. Args: n: Number of quaternions in a

Generate random rotations as 3x3 rotation matrices. Args: n: Number of rotation matrices in a batch to return. dtype: Type t

Read data from a pickle file.

Reveal some amino acids in the sequence according to schedule and predictions seq_t (torch.tensor): [L] Integer sequence se

reverse_sample uses an approximation to the IGSO3 score to sample a rotation at the previous time step. Roughly - this update

Creates initial features to start the sampling process.

Generate the next pose that the model should be supplied at timestep t-1.

sample_vec generates a rotation vector(s) from IGSO(3) at time steps ts. Return: Sampled vector of shape [len(ts), N, 3]

score_norm computes the score norm based on the time step and angle Args: t: integer time step omega: angles (scalar o

score_vec computes the score of the IGSO(3) density as a rotation vector. This score vecotr is in the direction of the sampled vector,

Calculates the index for discretized sigma during IGSO(3) initialization.

Perform spherical linear interpolation from the True coordinate frame for each residue to a randomly sampled coordinate frame

slerp_update uses SLERP to update the frames at time t to the predicted frame for t=0 Args: R_t, R_0: rotation matrices of shape [3,

Convert a unit quaternion to a standard form: one in which the real part is non negative. Args: quaternions: Quaternions with re

o3.spherical_harmonics returns different values on 1st call vs all subsequent calls All subsequent calls are reproducible.

Drops Hs from a dictionary like {'LG1': [('CB', 'C'), ('H2', 'H')]}

Serialize data into a pickle file.

convert cartesian coordinates into 2d distance and orientation maps Parameters ---------- xyz : pytorch tensor of shape [batch,n

convert template cartesian coordinates into 2d distance and orientation maps Parameters ---------- xyz_t : pytorch tensor of sha

Return the maximum number of timesteps that this design protocol will perform. Output: T (int):

Args: T: total number of time steps min_sigma: smallest allowed variance, should be at least 0.01 to maintain numeri

Parameters: truncation_level: for the igso3 numerical approximation

Parameters: noise_level: scaling on the noise added (set to 0 to use no noise, to 1 to have full noise)

Parameters: inputs: conf.scaffold_list as conf conf.inference.num_designs for sanity checking

Initialize sampler. Args: conf: Configuration.

Convert rotations given as axis/angle to rotation matrices. Args: axis_angle: Rotations given as a vector in axis angle form,

bin 2d distance and orientation maps

bin 2d distance and orientation maps

calc_score_norm_torch is a differentiable torch implementation of calc_score_norm Args: omega: length of an Euler vector (i.e. angle of r

calculate_igso3 pre-computes numerical approximations to the IGSO3 cdfs and score norms and expected squared score norms. Args: num_s

bin 2d distance maps

Convert rotations given as Euler angles in radians to rotation matrices. Args: euler_angles: Euler angles in radians as tensor of sh

Given currente chis, prediction of chi0, dynamic T, current t, and chi_beta_T schedules, sample new angles Need to

Given xt, predicted x0 and the timestep t, give mu of x(t-1) Assumes t is 0 indexed

Wrapper method for pulling an item from the list, and preparing ss and block adj features

Input: - idx: residue indices of given sequence (B,L) Output: - seqsep: sequence separation feature with sign (B, L, L, 1)

Returns t2d for a template. Parameters: xyz_t: [T, L, 36, 3] mask_t: [T, L, 36] seq_cat: [L] same_chain: [L,

Input: - xyz: current backbone cooordinates (B, L, 3, 3) - pair: pair features from Trunk (B, L, L, E) - idx: residue ind

Input: - xyz: current backbone cooordinates (B, L, 3, 3) - pair: pair features from Trunk (B, L, L, E) - idx: residue ind

Convert rotations given as rotation matrices to axis/angle. Args: matrix: Rotation matrices as tensor of shape (..., 3, 3). Ret

Convert rotations given as rotation matrices to Euler angles in radians. Args: matrix: Rotation matrices as tensor of shape (..., 3,

Converts rotation matrices to 6D rotation representation by Zhou et al. [1] by dropping the last row. Note that 6D representation is not uniq

Function to prepare inputs to diffusion model seq (torch.tensor, required): (L) integer sequence msa_masked (torch.ten

Apply the rotation given by a quaternion to a 3D point. Usual torch rules for broadcasting apply. Args: quaternion: Tensor of qu

Multiply two quaternions representing rotations, returning the quaternion representing their composition, i.e. the versor with nonnegative re

Generate a single random 3x3 rotation matrix. Args: dtype: Type to return device: Device of returned tensor. Default: if Non