Single-Molecule Peptide Discrimination via Flow-Through SERS and Machine Learning.

Area of Science:

• Biochemistry and Molecular Biology
• Analytical Chemistry
• Spectroscopy

Background:

• Peptides are crucial biomolecules in medicine, but their detection in complex mixtures is challenging.
• Surface-enhanced Raman spectroscopy (SERS) offers molecular specificity but faces limitations like spectral variability and overlap.
• Clinical applications require robust and accurate peptide detection methods.

Purpose of the Study:

• To investigate the performance of a SERS flow-through strategy using plasmonic nanopores for single-molecule peptide detection.
• To assess the feasibility of discriminating between closely related peptides using this technique.
• To establish flow-through nanopore SERS as a viable method for biomedical applications.

Main Methods:

• Utilized plasmonic nanopores to create sub-2 nm hotspots for SERS measurements.
• Employed electrophoretic delivery for controlled translocation of single peptide molecules.
• Recorded ultrafast SERS spectra on microsecond timescales using a single-photon avalanche diode camera.
• Applied machine-learning algorithms for spectral analysis and peptide classification.

Main Results:

• Achieved 70.5% classification accuracy for single-peptide discrimination (vasopressin vs. oxytocin).
• Improved discrimination accuracy to 99% by averaging spectra from 40 events.
• Demonstrated successful recording of Raman spectra from single molecules translocating through nanopores.

Conclusions:

• Flow-through nanopore SERS is a promising technique for single-molecule peptide identification.
• This method offers a potential solution for robust and accurate peptide detection in biomedical settings.
• The study highlights the capability of SERS and machine learning for differentiating structurally similar peptides.