Antimicrobial peptides (AMPs) are a promising new class of antibiotics, the mechanism of which usually involves binding to membranes with subsequent disruption. The disruption process is, however, still poorly understood, mostly because it seems to be mediated by states - pore formation or in-depth AMP insertion - that are unordered and very short-lived. My PhD work focused on using and developing biophysical methodologies to study the conditions at which the AMP-mediated disruption is triggered. Models bridging these results with in vivo observations were developed, establishing important relationships between AMP concentration and activity. Molecular Dynamics (MD) simulations allow the visualization of molecular level interactions with spatial and time resolution unsurpassed by experimental techniques, having already provided important clues on AMP-mediated membrane pore formation. In this project the interactions of AMPs with phospholipid bilayers and vesicles will be followed by MD simulations. A comparative study will be performed on a family of peptides derived from two naturally occurring AMPs (cecropin and melittin; base template XKLFKKILKXL-NH2) to gather insight on the different activities and toxicities of each of the peptides, and, from there, better understand the molecular characteristics responsible for those features. Work will also focus on the optimization and application of Coarse Grained (CG) simulation models, which allow significant reductions of processing time without major loss of detail. The interaction models that arise from MD simulations will be experimentally validated, making use of the researcher's background complemented by the host group's strong network of collaborating laboratories. The shift to computer simulation techniques will definitely broaden the researcher’s range of skills, with the added value that these techniques are easily portable and straightforward to implement independently in the future.