Random peptide libraries that cover large search spaces are often used for the discovery of new binders, even when the target is unknown. To ensure an accurate population representation, there is a tendency to use large libraries. However, parameters such as the synthesis scale, the number of library members, the sequence deconvolution and peptide structure elucidation, are challenging when increasing the library size. To tackle these challenges, we propose an algorithm-supported approach to peptide library design based on molecular mass and amino acid diversity. The aim is to simplify the tedious permutation identification in complex mixtures, when mass spectrometry is used, by avoiding mass redundancy. For this purpose, we applied multi (two- and three-)-objective genetic algorithms to discriminate between library members based on defined parameters. The optimizations led to diverse random libraries by maximizing the number of amino acid permutations and minimizing the mass and/or sequence overlapping. The algorithm-suggested designs offer to the user a choice of appropriate compromise solutions depending on the experimental needs. This implies that diversity rather than library size is the key element when designing peptide libraries for the discovery of potential novel biologically active peptides. Electronic supplementary material The online version of this article (10.1186/s13321-019-0347-6) contains supplementary material, which is available to authorized users.

No abstract

No abstract

The discovery of therapeutic peptides is often accelerated by means of virtual screening supported by machine learning-based predictive models. The predictive performance of such models is sensitive to the choice of data and its representation scheme. While the peptide physicochemical and compositional representations fail to distinguish sequence permutations, the amino acid arrangement within the sequence lacks the important information contained in physicochemical, conformational, topological, and geometrical properties. In this paper, we propose a solution to the identified information gap by implementing a hybrid scheme that complements the best traits from both approaches with the aim of predicting antimicrobial and antiviral activities based on experimental data from DRAMP 2.0, AVPdb, and Uniprot data repositories. Using the Friedman test of statistical significance, we compared our hybrid, sequential properties approach to peptide properties, one-hot vector encoding, and word embedding schemes in the 10-fold cross-validation setting, with respect to the F1 score, Matthews correlation coefficient, geometric mean, recall, and precision evaluation metrics. Moreover, the sequence modeling neural network was employed to gain insight into the synergic effect of both properties-and amino acid order-based predictions. The results suggest that sequential properties significantly (P < 0.01) surpasses the aforementioned state-of-the-art representation schemes. This makes it a strong candidate for increasing the predictive power of screening methods based on machine learning, applicable to any category of peptides.

No abstract