How to retrieve representative at 30% sequence identity in PDB

Question

[bosimiya]

I plan to use the PDB advanced search to filter sequences. I need to create a test set of protein sequences. The selection conditions are probably chain length, resolution, macromolecule type, etc., which are all easy to implement.

But there is another restriction: retrieving representative at 30% sequence identity. How do I achieve this?

Answer 1

[Mensur Dlakic]

First you download a FASTA file with all PDB sequences.

• https://ftp.wwpdb.org/pub/pdb/derived_data/pdb_seqres.txt.gz

Next you cluster them down to 30% identity using MMseqs2.