Single-Molecule Proteomics via a Dynamic Translocase and Physics-Informed Machine Learning

Single-molecule protein sequencing promises to revolutionize clinical proteomics, yet current platforms relying on static DNA-sequencing nanopores encounter a significant biophysical challenge. These "static calipers" are incapable of resolving isobaric residues, necessitating intricate DNA-handle chemistries and surpassing clinically relevant abundance levels for targets.

To overcome this limitation, researchers have developed a dynamic, target-docking translocase engine - the anthrax toxin protective antigen (PA) - as a label-free single-molecule peptide sensor. By analyzing the multi-state thermodynamic friction generated as the pore's active site dynamically adjusts around translocating analytes, a physics-informed machine learning (PIML) architecture was trained to classify a 20-member guest-host peptide library panel containing all 20 canonical amino acids at the single-event level.

Operating at low nanomolar concentrations within a 35-millisecond thermodynamic read constraint, this translocase successfully resolved isobaric variants, such as leucine and isoleucine. Moreover, a 98.02% (+/-0.05%) classification accuracy was achieved on a panel of five untagged, native clinical biomarkers, including KRAS G12D, angiotensin, and bradykinin.

By shifting from static volumetric measurement to time-domain thermodynamic fingerprinting, the researchers have established the necessary protein nanopore hardware for de novo proteomics.