(Nanocarbons Division Award, 2nd Place) Assessment of Biomolecule-Carbon Nanotube Interactions in Machine Perception Nanosensor Arrays

Arrays of chemically modified and DNA-wrapped carbon nanotube-based nanosensors are powerful new tools for biomarker-independent diagnostics. When exposed to patient biofluids, an individual nanosensor can simultaneously interact with many different biological molecules, enabling the detection of collective molecular changes, constituting a “spectral fingerprint”, rather than measuring individual biomarkers. Biomolecule-nanosensor interactions result in quantitative changes in sensor emission that are employed by machine learning models as features to differentiate between sample categories, such as healthy versus disease. To determine the underlying molecular interactions responsible for changes in sensor emission, we conducted single-analyte titrations of biomolecules at physiologically relevant concentrations. Each sensor exhibited optical responses to multiple biological molecules, illustrating the ability of the nanosensors to concurrently interact with many biomolecules to detect subtle molecular shifts occurring in different disease states.