protein sequences do not match the sequences in the pdb file

[Joe-Alwyn]

I found that in the protein.SHS27k.sequences.dictionary.tsv file, some of the protein sequences do not match the sequences in the pdb file. Could you please clarify which one should be taken as the standard?

[LirongWu]

We used AlphaFold2 to predict the 3D structure (PDB) of the sequence, which might have introduced some inconsistencies; in case of conflicts, we recommend using the PDB results as the reference. For your convenience, you may directly use our pre-processed files.