Opto-acoustic sensor for protein conformational fingerprinting beyond concentration degeneracy: analytical modelling and AI-based sorting.

We propose a dual-observable sensor that resolves the concentration-conformation degeneracy of existing biosensors by combining whispering-gallery-mode lasing spectroscopy with acoustic radiation force modulation in a microdroplet. When an acoustic field is applied, proteins concentrate at a rate governed by acoustic contrast features. Upon field removal, they recover at a rate that reflects their hydrodynamic radius through the Stokes-Einstein relation. These time-resolved observables form a two-dimensional conformational fingerprint. Through systematic study, Monte Carlo robustness analysis, and SVM classification, conformational states are discriminated with greater than 99% accuracy. The platform requires no labelling and prior concentration information, offering a novel path toward quantitative protein sensing.