Why not just get the protein sequences and write or ask an LLM to provide an awk or python script to count the number of cysteines in the sequence.

You could optionally normalize the number of Cs by the protein sequence length to get %Cs.

There is a subset of BioPython called SeqIO with many examples of how to read and manipulate biological sequences. You would simply need to add a cysteine counter. Here is a basic script:

import sys
from Bio import SeqIO

FastaFile = open(sys.argv[1], 'r')

for rec in SeqIO.parse(FastaFile, 'fasta'):
    name = rec.id
    seq = str(rec.seq)
    cys = seq.lower().count('c')
    print(name, len(seq), ('%.4f' % float(cys/len(seq))))

FastaFile.close()

This will print each sequence name, total length and a fraction of cysteines. It is easy to customize it so only sequences with higher fraction of cysteines are listed.