|About this Abstract
||2018 TMS Annual Meeting & Exhibition
||Bio-nano Interfaces and Engineering Applications Symposium
||New Antimicrobial Peptides Generated through Genetic Algorithm Approach Using Chemical Property Based Cross-over
||Kyle Boone, Kyle Camarda, Paulette Spencer, Candan Tamerler
|On-Site Speaker (Planned)
Antibiotic-resistant bacteria cause many fatalities annually. Antimicrobial peptides have co-evolved with bacteria as natural antibiotics. We previously demonstrated that antimicrobial peptides when combined with an implant binding peptide can create a biomimetic interface and coat the implant surfaces. Motivated by these results, we also developed successful searches for new antimicrobial peptides using a genetic algorithm (GA) method. To target specific bacteria as well as broaden their effect, here we explore developing new sequences using computer-aided molecular design (CAMD) principles to translate the desired properties into novel antimicrobial peptide sequences. We advance our GA method by developing a novel sequence generation mechanism. We introduce the chemical property based cross over concept by grouping sequences before sequence generatation. The convergence rate of the average distance between property values and targets across generations will be compared between using inter-group crosses, intra-group crosses and both types of crosses to generate new sequences.
||Planned: Publication Outside of TMS (Indicate publication title and publisher if known.)