Detect and verify JAX GPU device availability. Checks for JAX GPU devices and returns available GPU device list. Provides comprehensive G

Get all residues within a specified distance of a target residue Args: pose: PyRosetta Pose object target_residue_number: in

Returns the oldest revision date.

get sidechain indices/weights for all_atom14_positions

get sequences as strings - set get_best=False, to get the last sampled sequence

Extract protein sequences from PDB/mmCIF structure files. Parses structural files to extract amino acid sequences for each protein chain.

Detect and return available PyTorch device. Checks for CUDA availability and returns the appropriate device string for PyTorch operations

Consolidate all trajectory information into a single dictionary. Combines trajectory metrics, filtering metrics, and other metrics wi

Returns identity of given shape.

Convert range string specification to zero-based index list. Parses a range string (e.g., 'B1-10,A5') into a list of zero-based indices,

Encodes a number as a string, using reverse spreadsheet style naming. Args: num: A positive integer. Returns: A string that encodes the

Computes inverse of rotations 'm'.

Check whether the binder interface is near target hotspot residues. Args: target_contacts: Target residue indices at the interface (1-ind

Load stereo_chemical_props.txt into a nice structure. Load literature values for bond lengths and bond angles and translate bond angles into the

Log design trajectory metrics to CSV file for analysis. Extracts and saves various optimization metrics from the design trajectory to a C

Returns translation and rotation matrices to canonicalize residue atoms. Note that this method does not take care of symmetries. If you provide the

Solves multi-task gradient combination using quadratic programming with regularization. Implements the Multi-Task Learning as Multi-Objective

Extracts loop associated with a prefix from mmCIF data as a dictionary. Args: prefix: Prefix shared by each of the data items in the loop.

input: x = PDB filename atoms = atoms to extract (optional) output: (length, atoms, coords=(x,y,z)), sequence

input: x = PDB filename atoms = atoms to extract (optional) output: (length, atoms, coords=(x,y,z)), sequence

Get the atm record

Parses FASTA string and returns list of strings with amino-acid sequences. Arguments: fasta_string: The string contents of a FASTA file. Ret

PCGrad algorithm for two gradients using numpy. Args: grad1 (np.ndarray): First gradient vector grad2 (np.ndarray): Seco

Read a pdb file predicted with AF and rewritten to conatin all chains

read pdb file and return as string

use py3Dmol to plot pdb coordinates - color=["pLDDT","chain","rainbow"]

Generate and save trajectory loss plots for design analysis. Creates individual plots for key optimization metrics throughout the design

Return a new QuatAffine which applies the transformation update first. Args: update: Length-6 vector. 3-vector of x, y, and z such that the

given [L]ength, [N]umber of sequences and number of [T]emplates return dictionary of blank features

Runs alignment tools on the input sequences and creates features.

Create pseudo beta features.

single sequence

Multiply a quaternion by a pure-vector quaternion.

Convert a normalized quaternion to a rotation matrix.

Change the design name identifier for this trajectory. Args: new_name (str): New unique identifier for this trajectory.

restart the optimization ------------ note: model.restart() resets the [opt]ions and weights to their defaults use model.set_opt(...,

Given a list of segment [length]s, move them around given an [offset], [order] and [loop] lengths. The [order] of the segments and the length of

Run AF3 or Chai structure prediction for antibody-target complex. Args: trajectory_sequence: Designed antibody sequence targ

sample sequence

sample new sequence(s) in parallel

save pdb coordinates (if filename provided, otherwise return as string)

Save run and target configuration to persistent YAML file. Persists the complete configuration for the design run to enable r

Scale translation in Rigid3Array by 'factor'.

I = { [[required]] 'X' = (L,4,3) 'mask' = (L,) 'residue_index' = (L,) 'chain_idx' = (L,)

Computes square of euclidean distance between 'vec1' and 'vec2'. Args: vec1: Vec3Array to compute distance to vec2: Vec3Array to compute

do one step of gradient descent

Make layer symmetric in final two dimensions, used for contact prediction.

Embed templates into an (num_res, num_templates, num_channels) embedding. Args: batch: A batch containing: template_aatype, (num_template

given coordinates a-b-c, return angle

convert to float

given coordinates a-b, return length or distance

Update core trajectory optimization metrics. Updates the trajectory metrics with values from the design optimization process.