Computational protein design can create binders against targets of interest, but identifying binders with sufficient affinity still requires biochemical screening of many designs. In this work, we test flow-induced dispersion analysis (FIDA) as a method for screening binders in a time and cost-effective manner. FIDA uses Taylor dispersion analysis to determine the hydrodynamic radius of fluorescently labeled biomolecules and their complexes. Here, we use FIDA to assess the binding of RFdiffusion-designed protein binders against the small helical peptide ALFA-tag and the guanylate kinase (GK) domain of PSD-95. Successful binders can be identified in a single measurement using heat-treated bacterial lysates, allowing rapid identification of binders with high affinity and thermostability. Subsequent titration experiments show micromolar affinities for ALFA-tag binders and nanomolar affinities for GK domain binders. The lack of immobilization, the minimal sample volume, and the compatibility with complex biological samples position FIDA as a valuable tool for the screening and characterization of computationally generated protein binders.
Pinheiro et al. (Wed,) studied this question.