mindsponge.common.find_optimal_renaming

View Source On Gitee
mindsponge.common.find_optimal_renaming(atom14_gt_positions, atom14_alt_gt_positions, atom14_atom_is_ambiguous, atom14_gt_exists, atom14_pred_positions)[source]

Find optimal renaming for ground truth that maximizes LDDT.

Reference:

Jumper et al. (2021) Suppl. Alg. 26 "renameSymmetricGroundTruthAtoms"

Parameters

  • atom14_gt_positions (Tensor) – Ground truth positions in global frame with shape \((N_{res}, 14, 3)\).

  • atom14_alt_gt_positions (Tensor) – Alternate ground truth positions in global frame with coordinates of ambiguous atoms swapped relative to 'atom14_gt_positions'. The shape is \((N_{res}, 14, 3)\).

  • atom14_atom_is_ambiguous (Tensor) – Mask denoting whether atom is among ambiguous atoms, see Jumper et al. (2021) Suppl. Table 3. The shape is \((N_{res}, 14)\).

  • atom14_gt_exists (Tensor) – Mask denoting whether atom at positions exists in ground truth with shape \((N_{res}, 14)\).

  • atom14_pred_positions (Tensor) – Predicted positions of atoms in global prediction frame with shape \((N_{res}, 14, 3)\).

Returns

Tensor, \((N_{res},)\) with 1.0 where atom14_alt_gt_positions is closer to prediction and otherwise 0.

Supported Platforms:

Ascend GPU

Examples

>>> import numpy as np
>>> from mindsponge.common.utils import find_optimal_renaming
>>> from mindspore import Tensor
>>> n_res = 16
>>> atom14_gt_positions = Tensor(np.random.randn(n_res, 14, 3).astype(np.float32))
>>> atom14_alt_gt_positions = Tensor(np.random.randn(n_res, 14, 3).astype(np.float32))
>>> atom14_atom_is_ambiguous = Tensor(np.random.randn(n_res, 14).astype(np.float32))
>>> atom14_gt_exists = Tensor(np.random.randn(n_res, 14).astype(np.float32))
>>> atom14_pred_positions = Tensor(np.random.randn(n_res, 14, 3).astype(np.float32))
>>> out = find_optimal_renaming(atom14_gt_positions, atom14_alt_gt_positions,
...                             atom14_atom_is_ambiguous, atom14_gt_exists, atom14_pred_positions)
>>> print(out.shape)
(16,)