The evaluation results are quite different from those in the paper.

[mapengsen]

I downloaded the AlphaFlow-MD base dataset provided on GitHub, and then compared it with the files downloaded by download_atlas.sh. I found that the results were extremely different. Could it be that my evaluation was incorrect? My comparison script is:

python -m scripts.analyze_ensembles --atlas_dir /mnt/f/AAApublish_paper_code_paper/doctor/MD_agent/alphaflow/download_MD --pdbdir /mnt/f/AAApublish_paper_code_paper/doctor/MD_agent/alphaflow/esmflow_md_base_202402 --num_workers 10

python -m scripts.print_analysis out.pkl