Download model/PXDesignBench/pxdbench/metrics/diversity.py from OneScience-Group/PXDesign: direct link, hf CLI and curl.
- Browser
- Download file 2.06 kB
-
https://huggingface.co/OneScience-Group/PXDesign/resolve/main/model/PXDesignBench/pxdbench/metrics/diversity.py
- Command line
-
hf download hf://OneScience-Group/PXDesign/model/PXDesignBench/pxdbench/metrics/diversity.py
-
curl -L -o diversity.py https://huggingface.co/OneScience-Group/PXDesign/resolve/main/model/PXDesignBench/pxdbench/metrics/diversity.py
2.06 kB
| # Copyright 2025 ByteDance and/or its affiliates. | |
| # | |
| # Licensed under the Apache License, Version 2.0 (the "License"); | |
| # you may not use this file except in compliance with the License. | |
| # You may obtain a copy of the License at | |
| # | |
| # http://www.apache.org/licenses/LICENSE-2.0 | |
| # | |
| # Unless required by applicable law or agreed to in writing, software | |
| # distributed under the License is distributed on an "AS IS" BASIS, | |
| # WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. | |
| # See the License for the specific language governing permissions and | |
| # limitations under the License. | |
| import multiprocessing as mp | |
| from itertools import combinations | |
| from pxdbench.metrics import tmalign | |
| def compute_pair_tm_score(args): | |
| file1, file2, chain_id = args | |
| tm_score = tmalign.get_tm_score(file1, file2, chain_id, chain_id) | |
| return tm_score | |
| def compute_diversity(pdb_files, chain_id=None): | |
| """ | |
| Calculate the average TM-score between all pairs of PDB files to measure diversity. | |
| Args: | |
| pdb_files (list): A list of file paths to PDB files. | |
| chain_id (str, optional): The chain ID to use for TM-score calculation. Defaults to None. | |
| Returns: | |
| float or None: The average TM-score between all valid pairs of PDB files. | |
| Returns None if no valid pairs are found. | |
| """ | |
| num_workers = min(32, mp.cpu_count()) | |
| pairs = list(combinations(pdb_files, 2)) | |
| with mp.Pool(processes=num_workers) as pool: | |
| args_list = [(file1, file2, chain_id) for file1, file2 in pairs] | |
| tm_scores = pool.map(compute_pair_tm_score, args_list) | |
| valid_scores = [score for score in tm_scores if score is not None] | |
| count = len(valid_scores) | |
| sum_scores = sum(valid_scores) | |
| if count > 0: | |
| average = sum_scores / count | |
| print(f"avg pair TMscore: {average:.4f}") | |
| print(f"total nums pair: {count}") | |
| else: | |
| average = None | |
| print("No valid TM-score pairs found.") | |
| return average | |