Adeno-associated virus characterization for cargo discrimination through nanopore responsiveness

Area of Science:

• Nanotechnology
• Virology
• Biophysics

Background:

• Solid-state nanopore (SSN) technology is increasingly used in genomics and proteomics.
• Its application in virology, crucial for infectious disease and drug delivery research, is emerging.
• Accurate characterization of viral vectors like adeno-associated viruses (AAVs) is vital for clinical applications.

Purpose of the Study:

• To demonstrate SSN's capability in discriminating AAVs based on their genetic payload (double-stranded DNA, single-stranded DNA, or empty).
• To develop a machine learning model for accurate classification of AAVs using SSN data.
• To assess the potential of SSN for detecting empty AAVs in mixtures with DNA-containing AAVs.

Main Methods:

• Utilized SSN to measure nanopore-induced electro-deformation (relative current change, ΔI/I0) of AAVs.
• Developed a hybrid deep learning algorithm combining support vector machine and neural network.
• Analyzed resistive-pulse signals generated by AAVs passing through the nanopore.

Main Results:

• A distinct electro-deformation order was observed: AAVempty > AAVssDNA > AAVdsDNA.
• The deep learning algorithm achieved >95% accuracy in classifying AAV types based on their genetic cargo.
• The method successfully flagged the presence of AAVempty in mixtures with AAVdsDNA at varying concentrations.

Conclusions:

• SSN analysis, without prior digestion, effectively differentiates AAVs by their genetic content.
• The developed deep learning approach offers a highly accurate tool for AAV classification.
• This SSN-based method shows promise for clinical diagnostics and biomedical research involving viral vectors.