I wonder how to modify the known antibacterial peptides’ amino acids composition to make them more easily bond to both gram- positive and negative bacteria, which kinds of aa replaced could enhance this ability ?
Generally you will want your peptides to possess an amphipathic tertirary structure. You will want to have a large number of hydrophobic amino acids, and a number of positively charged residues.
Avoid negatively charged residues. A number of prediction services exist for prediction of anitimicrobial peptide activity, you can use those to evaluate your substitutions.