Reducing hurdles to clinical trials without compromising the therapeutic promises of peptide candidates becomes an essential step in peptide-based drug design. Machine-learning models are cost-effective and time-saving strategies used to predict biological activities from primary sequences. Their limitations lie in the diversity of peptide sequences and biological information within these models. Additional outlier detection methods are needed to set the boundaries for reliable predictions; the applicability domain. Antimicrobial peptides (AMPs) constitute an extensive library of peptides offering promising avenues against antibiotic-resistant infections. Most AMPs present in clinical trials are administrated topically due to their hemolytic toxicity. Here we developed machine learning models and outlier detection methods that ensure robust predictions for the discovery of AMPs and the design of novel peptides with reduced hemolytic activity. Our best models, gradient boosting classifiers, predicted the hemolytic nature from any peptide sequence with 95–97% accuracy. Nearly 70% of AMPs were predicted as hemolytic peptides. Applying multivariate outlier detection models, we found that 273 AMPs (~ 9%) could not be predicted reliably. Our combined approach led to the discovery of 34 high-confidence non-hemolytic natural AMPs, the de novo design of 507 non-hemolytic peptides, and the guidelines for non-hemolytic peptide design.

Current cancer treatments damage healthy cells and tissues, causing short‐term and long‐term side effects. New treatments are desired that show greater selectivity toward cancer cells and evade the common mechanisms of multidrug resistance. Membranolytic anticancer peptides (mACPs) hold promise against cancer and multidrug resistance. Amphipathicity, hydrophobicity, and net charge of mACPs participate in their respective interactions with cell membranes and their overall inhibition of cancer cells. To support the design of cell‐line selective mACPs, we investigated the relationships that amino acid composition, physicochemical properties, sequence motifs, and sequence homology could have with their potency and selectivity towards several healthy and cancer cell lines. Sequence length and net charge are known to affect the selectivity of mACPs between cancer and healthy cell lines. Our study reveals that increasing the net charge or flexibility (i. e., small and aliphatic residues) influences their selectivity between cancer cell lines with comparable lipid compositions.

Current cancer treatments damage healthy cells and tissues, causing short-term and long-term side effects. New treatments are desired that show greater selectivity toward cancer cells and evade the common mechanisms of multidrug resistance. Membranolytic anticancer peptides (mACPs) hold promise against cancer and multidrug resistance. The amphipathicity, hydrophobicity, and net charge of mACPs are all known to play a role in their respective interactions with cell membranes and the overall biological inhibition of cancer cells, but they are insufficient to explain their cell-line selectivity. To support the design of cell-line selective mACPs, we investigated the relationships that peptide features (amino acid composition, physicochemical properties, sequence motifs and sequence homology) could have with their potency and selectivity towards several healthy and cancer cell lines. Hydrophobicity, net charge and the presence of small and aliphatic residues influenced the potency and selectivity of mACPs across cancer cell lines in a cell-specific manner.