Bottom-Up Design Approach for OBOC Peptide Libraries

Abstract

One-bead-one-compound peptide libraries, developed following the top-down experimental approach, have attracted great interest in the identification of potential ligands or active peptides. By exploiting a reverse experimental design approach based on the bottom-up strategy, we aimed to develop simplified, maximally diverse peptide libraries that resulted in the successful characterization of mixture components. We show that libraries of 32 and 48 components can be successfully detected in a single run using chromatography coupled to mass spectrometry (UPLC-MS). The proposed libraries were further theoretically evaluated in terms of their composition and physico-chemical properties. By combining the knowledge obtained on single libraries we can cover larger sequence spaces and provide a controlled exploration of the peptide chemical space both theoretically and experimentally. Designing libraries by using the bottom-up approach opens up the possibility of rationally fine-tuning the library complexity based on the available analytical methods.

Esterase Sequence Composition Patterns for the Identification of Catalytic Triad Microenvironment Motifs

“…The R package Peptides was used to calculate the aa composition of peptides, expressed in mole percentage (mol%). The aa are categorized based on their intrinsic properties into 11 subcategories: Tiny (Ala, Cys, Gly, Ser, Thr), Small (Ala, Cys, Asp, Gly, Asn, Pro, Ser, Thr, Val), Aliphatic (Ala, Ile, Leu, Val), Aromatic (Phe, His, Trp, Tyr), Nonpolar (Ala, Cys, Phe, Gly, Ile, Leu, Met, Pro, Val, Trp, Tyr), Polar (Asp, Glu, His, Lys, Asn, Gln, Arg, Ser, Thr), Charged (Asp, Glu, His, Lys, Arg), Basic (His, Lys, Arg), Acidic (Asp, Glu), Sulfur (Cys, Met), and Hydroxylic (Ser, Thr).…”

Designed Complex Peptide‐Based Adaptive Systems: A Bottom‐Up Approach

“…Their method showed that genetic algorithms can optimize the library search with the highest number of peptides based on multiple objectives such as chemical and mass diversity. Their system could be extended to other search objectives and their combinations as they've later shown by introducing a sequence‐based property evaluation step, such as polarity, hydrophobicity and/or H‐bonding [36] . Their work opens up the possibilities of user‐defined, rationally designed peptide libraries by exploring the sequence space in an informed and targeted way.…”

Sequential Properties Representation Scheme for Recurrent Neural Network-Based Prediction of Therapeutic Peptides

“…12,13 By incorporating the diversity parameter into the evolutionary process, we emphasized the intra-library diversity, which makes individual peptides distinguishable by mass spectrometry. 14 The peptides were represented as strings, i.e., sequences of amino acid single letter codes, matching the FASTA format. Novel ML methods, mainly generative adversarial networks and variational autoencoders, present an alternative to genetic algorithms as a means of generating novel peptides exhibiting specific activity.…”

“…Genetic algorithms have been used to construct therapeutic, bactericidal, and antimicrobial peptides. − In our previous work, we have presented a genetic algorithm adaptation and parametrization for exploring the chemical space of short peptides using random peptide libraries. , By incorporating the diversity parameter into the evolutionary process, we emphasized the intra-library diversity, which makes individual peptides distinguishable by mass spectrometry . The peptides were represented as strings, i.e., sequences of amino acid single letter codes, matching the FASTA format.…”