Design and Optimization of a Novel Antimicrobial Peptide (AMP) to Inhibit Klebsiella Pnuemoniae

Antibiotic resistance in Klebsiella pneumoniae is an urgent global health challenge, which is driven by widespread multidrug resistance and the emergence of carbapenem-resistant strains. Alternatively, antimicrobial peptides (AMPs) represent a promising alternative therapeutic strategy due to their ability to disrupt bacterial membranes and reduce the likelihood of resistance development. This study aimed to design and computationally optimize a peptide with increased predicted AMP probability using bioinformatic tools. The initial sequence that was started with was KRFHSVGSLIQRLQQMIRDLSEATRHGIRKITRPKLLLAS, which scored only 0.534 on the CAMPr3 Random Forest classifier. Through iterative design steps, including shortening the sequence, increasing net positive charge, balancing hydrophobicity, and optimizing residue spacing, the final sequence, LIQRLAKMIRKLSHAIRHWIRKI, with an AMP probability score of 0.9855, a net charge of +7.5, and hydrophobic content of 50% was developed. Structural modeling in AlphaFold confirmed an amphipathic α-helical shape, and ToxinPred predicted non-toxicity of the peptide sequence. These results demonstrate how computational methods can accelerate peptide design and generate strong therapeutic candidates for combating multidrug resistant pathogens.