The mechanism by which cell-penetrating peptides and antimicrobial peptides cross plasma membranes is unknown, as is how cell-penetrating peptides facilitate drug delivery, mediating the transport of small molecules. Once non-disruptive and non-endocytotic pathways are excluded, pore formation is one of the proposed mechanisms, including toroidal, barrel-stave, or carpet models.Spontaneous pores have been observed in coarse-grained simulations and less often in molecular dynamics simulations. While pores are widely assumed and inferred, there is no unambiguous Antibiotic-resistant bacteria constitute a significant public health problem 1,2 . Antimicrobial peptides (AMPs) -a class of small (≤50 residues), cationic or amphiphilic compounds -are amongst the most promising solutions currently known 3 . Cell-penetrating peptides (CPPs) are a similar class of peptides, important for their binding -usually via covalent conjugation through a disulfide bridge or by means of non-covalent interactions 4 -and delivery of drugs into cells for therapeutic applications [5][6][7] . Both CPPs and AMPs interact strongly with the membrane surface, penetrating or altering its permeability via a mechanism which is not well understood 8,9 , even after a plethora of experimental and simulation studies.Atomistic molecular dynamics (MD) can be applied to investigate certain aspects of the uptake mechanisms. Despite its limitations in terms of accessible spatial and temporal scales, MD has the potential to provide unique insights. Several studies have converged, via observations or deductions, to infer the formation or existence of pores in the membrane 10-14 . Pores have been observed in coarse-grained simulations 15,16 , and in molecular dynamics 17,18 . A spontaneous translocation of a TAT peptide was observed in one MD study 19 and explained in terms of pore formation. However, subsequently this result was found to have originated from incorrect use of electrostatic interactions 14 . When a phosphatidylcholine (POPC) membrane is exposed to a CPP such as maculatin 1.1, dye leakage was observed after ∼10 minutes. This result led the authors of the study to infer a physical mechanism based on membrane disruption or pore formation 20 .Binding energies of pore formation with alamethicin and melittin showed preference for cylindrical 21 and toroidal 22 pores, in accord with molecular dynamics results 23 .In contrast to the above picture, some experiments do not support the presence of pores.Considering that cationic peptides should interact more strongly with negatively charged compounds, one would have expected pore formations with a CPP oligoarginine peptide interacting with negatively charged lipids. Interestingly, a leakage was observed, but not with an ion-conductive structure 24 , which would have been expected in presence of holes or channels.Trans-activating transcriptional (TAT) peptide penetration has been observed in the absence of endocytosis, with the additional discovery that added fluorophores were not taken together with the...