Surface-enhanced Raman spectroscopy (SERS)-enabled detection of nucleic acid
Machine learning application towards spectroscopy-based environmental dataset
Rapid nucleic acid analysis is of great interest for the characterization of microbial communities and biological functions in molecular biology and genetics across numerous science, engineering, and industrial sectors. Surface-enhanced Raman spectroscopy (SERS) is a promising candidate as an analytical bio-sensing tool for nucleic acid detection. Inelastic Raman scattering that arises from vibrational modes within molecules provides unique molecular fingerprints. The application of Raman scattering for sensing has historically been limited due to the intrinsic small Raman cross-section of molecules. Following the discovery of SERS, the phenomenon whereby the Raman signal is significantly enhanced by a factor of 105-106 when the molecule of interest is situated adjacent to the surface of a plasmonic metal substrate, interest in Raman based methods for chemical and biological analysis has been rapidly increasing.
To achieve higher discriminatory capacity between SERS spectra of gene sequences, more powerful classifiers are required. Herein, we built a flexible discriminatory tool by combining a tree-based decision rule and a multiclass support vector machine (Tr-SVM) for the identification of gene sequences based on their SERS spectra. A tree-based decision rule can group the correlated classes and provide multiple classifiers based on the decision levels. Each classifier can be better optimized to the dataset of each classified group than one classifier. SVM is a machine learning technique that differentiates multi-dimensional data by a separating hyperplane. The optimal hyperplane maximizes the margin of the data from different classes.
Synthesis of the novel nanocomposites for SERS application
Bacterial cellulose nanocrystals (BCNCs) are biocompatible cellulose nanomaterials that can host guest nanoparticles to form hybrid nanocomposites with a wide range of applications. Herein, we report the synthesis of gold (AuNP) and iron oxide (Fe3O4) nanoparticle functionalized BCNCs (Au@Fe3O4@BCNCs) that combine the magnetic separation enabled by Fe3O4 and the plasmonic properties of AuNPs.