Assuming I already have a structure of a receptor bound to a peptide and I want to identify base pair substitutions in the peptide that can increase its affinity.
There are many criteria according to which a peptide could be modified in order to optimize its bonding to a receptor's site. Among them, just to mention some, are hydrophobicity, bulkiness, antigenicity, ... . However, I think the most promising one is the antisense-peptides strategy that offers the advantage of selectivity as an extra.
There were plenty of computer algorithms at 80-90s, when I was still active on these.