Diversification in Genetically Encoded Display

Directed evolution is a robust strategy for rapidly evolving genes, enabling them to adapt under specific selection pressures. Peptide and protein display is a powerful example of this approach, linking genotype to phenotype and then subjecting expressed sequences to selective forces to generate new function in the form of targeting binding. In recent decades, display techniques have primarily been employed to identify peptide and protein binders against specific targets. Although proteinogenic amino acids offer a diverse range of chemical and physical properties, their limitation in number can pose challenges—such as the lack of functional groups necessary for targeting certain pharmacophores, susceptibility to proteolytic degradation, and poor cell membrane permeability. To overcome these limitations, various strategies have been developed that incorporate non-proteinogenic moieties through ribosomal incorporation of non-proteinogenic amino acids (npAAs) and/or post-translational modifications of both the proteinogenic and non proteinogenic amino acids. These approaches have enabled the introduction of novel chemical functionalities, macrocyclization of peptides, and the attachment of additional fragments, thereby broadening the spectrum of accessible targets. This thesis leverages chemical diversification strategies in peptide display technology. Specifically, we intensively review the existing diversification available in peptide display, followed by the comparison with DNA-encoded libraries and its potential application in combination with mRNA display technique. To grasp the boundary of chemical space in mRNA display, we investigated the versatility of mRNA display across a range of chemical reactions. By employing selected reactions, we demonstrated the feasibility of monosaccharide-modified peptide displays—a modification that, while understudied, holds significant potential for expanding the chemical space accessible through peptide display. In summary, the thesis not only defines the chemical and physical parameters of mRNA display, but also expands its utility by post translational modification beyond traditional limits. By examining and incorporating diverse chemical modification conditions, this work paved the way for screening a broader class of molecules such as, but not limited to, glycopeptides. These additional diversification approaches would enable targeting of more challenging molecules in the future. Additionally, the advanced sequence analysis introduced in this thesis provides a robust framework for comprehensive sequence mining. This analysis workflow enabled efficient exploration of sequence outputs and revealed information, which was concealed by artificial thresholds. As more developed sequencing becomes available, the demand for equally sophisticated analytical tools will intensify. In this context, the growing development of digital technologies, including artificial intelligence, holds tremendous promise for data processing, sequence pattern identification, and binding prediction without human biases. We foresee a transformative synergy between these digital advancements and high throughput screening thanks to the huge number of datapoints and already existing data for training algorithms. Together, the innovations will chart a path toward a future, where molecular discoveries is not just faster, broader, and smarter but boldly reimagined.