New method driven by machine learning allows for the accurate measurement of the mass of individual particles and molecules using complex nanoscale devices, opening possibilities for the identification of proteins and potentially determining the sequence of the complete proteome. Proteins play a crucial role in living systems, and understanding which proteins are present, where they are produced, and in what quantities can offer valuable insights into health, disease, and strategies for combating illnesses. However, characterizing entire proteomes has been a challenge for scientists.
Mass spectrometry is a popular analytical tool used in various molecular investigations, involving ionizing samples, sending them on a specific path, and observing how they move towards a magnetic or electric field to determine mass and charge information. While mass spectrometry has been utilized for multiple purposes, biological samples can be challenging due to potential alterations from the ionization process. Developing sophisticated nanoscale devices called nanoelectromechanical systems (NEMS) has enabled the measurement of the masses of small molecules in real-time without the need for initial ionization.
NEMS devices involve a small beam that resonates when struck, providing information on the vibrational modes and frequency changes when a sample is placed on it. However, variations or imperfections at the nanoscale level can impact device-to-device consistency, resulting in challenges in determining precise mode shapes. The location of the sample within the device can also affect frequency measurements, complicating the process further. Caltech researchers have introduced a new technique called “fingerprint nanoelectromechanical mass spectrometry” to address these issues by measuring frequency changes in real-time, building a library of vectors for machine-learning software based on unique fingerprints.
Each vector represents a fingerprint that changes uniquely depending on where a particle lands on the NEMS device, offering a direction that facilitates comparison with the database for determining an unknown particle’s mass. The fingerprint technique can be utilized with any device, even advanced NEMS devices that trap vibrations for extended periods, such as phononic crystal NEMS devices. The team used alternate devices to benchmark the fingerprint method, including measuring the mass of individual particles of GroEL, a protein involved in protein folding within cells. Traditional mass spectrometry methods are challenging for large protein complexes and membrane proteins like GroEL, requiring sample fragmentation for analysis.
The new fingerprint technique aims to move towards native single-molecule mass spectrometry, enabling the examination of large proteins and protein complexes in their native form without fragmentation. This innovative approach could revolutionize proteomics by providing a means to analyze proteins one by one, gaining insights into their configurations and conformations without destroying them in the process. With its potential to measure millions of proteins within a reasonable timeframe and understand complete proteomes, the fingerprint nanoelectromechanical mass spectrometry method offers an exciting opportunity in the field of molecular analysis and could lead to significant advancements in health and disease research.