Evaluation of FRET X for single-molecule protein fingerprinting

Single-molecule protein identification is an unrealized concept with potentially ground-breaking applications in biological research. We propose a method called FRET X (Forster Resonance Energy Transfer via DNA eXchange) fingerprinting, in which the FRET efficiency is read out between exchangeable dyes on protein-bound DNA docking strands and accumulated FRET efficiencies constitute the fingerprint for a protein. To evaluate the feasibility of this approach, we simulated fingerprints for hundreds of proteins using a coarse-grained lattice model and experimentally demonstrated FRET X fingerprinting on model peptides. Measured fingerprints are in agreement with our simulations, corroborating the validity of our modeling approach. In a simulated complex mixture of >300 human proteins of which only cysteines, lysines, and arginines were labeled, a support vector machine was able to identify constituents with 95% accuracy. We anticipate that our FRET X fingerprinting approach will form the basis of an analysis tool for targeted proteomics.

Automated summary

The study introduces a method called FRET X fingerprinting for single-molecule protein identification, which aims to create fingerprints for proteins. Through simulations and experimental validation, the effectiveness of the method was demonstrated, successfully identifying complex mixtures including multiple proteins, providing a new analysis tool for targeted proteomics.