Summary Phenotypic drug discovery (PDD) enables the target-agnostic generation of therapeutic drugs with novel mechanisms of action. However, realizing its full potential for biologics discovery requires new technologies to produce antibodies to all, a priori unknown, disease-associated biomolecules. We present a methodology that helps achieve this by integrating computational modeling, differential antibody display selection, and massive parallel sequencing. The method uses the law of mass action-based computational modeling to optimize antibody display selection and, by matching computationally modeled and experimentally selected sequence enrichment profiles, predict which antibody sequences encode specificity for disease-associated biomolecules. Applied to a phage display antibody library and cell-based antibody selection, ∼10 5 antibody sequences encoding specificity for tumor cell surface receptors expressed at 10 3 –10 6 receptors/cell were discovered. We anticipate that this approach will be broadly applicable to molecular libraries coupling genotype to phenotype and to the screening of complex antigen populations for identification of antibodies to unknown disease-associated targets. Graphical abstract Highlights • Generation of antibodies based on in silico -predicted antibody enrichment signatures • Integrates computational modeling, differential antibody display selection, and NGS • Optimizes antibody display selection by in silico modeling • Generates a diverse antibody pool targeting a broad range of surface antigens Motivation Phenotypic antibody discovery enables the identification of novel antibodies with the most promising functional in vitro or in vivo activity, without prior knowledge of the targeted antigen. For efficient phenotypic discovery, a large panel of antibodies against a broad repertoire of potential targets should be included in the functional testing. Current methods generate limited numbers of antibodies (10 2 –10 3 ), targeting a few highly expressed antigens. This is a concern because antibodies to low-expressed antigens may have functional activity and may be relevant to biomarker discovery and therapeutic antibody development. We present a methodology that significantly enhances the number of antibodies generated (10 5 ) and expands the antibody-targeted receptor expression range (antigens differentially expressed at 10 3 –10 6 copies/cell) to include low-expressed tumor-selective antigens, enabling functional testing of a large pool of antibodies targeting a broad range of surface antigens. Phenotypic antibody discovery can identify antibodies with novel mechanisms of action, but it suffers from shortcomings in generating diverse antibody pools for functional testing. Mattsson et al. integrate computational modeling, differential antibody display selection, and massively parallel sequencing to generate diverse antibody pools to the cell surfaceome.
【저자키워드】 Biomarkers, Computational biology, mathematical modeling, phage display, therapeutic antibodies, phenotypic antibody discovery, specificity predictions,