SERS and advanced chemometrics – Utilization of Siamese neural network for picomolar identification of beta-lactam antibiotics resistance gene fragment

Temperature-Responsive Nanocarrier-Regulated Alternative Release of “Cargos” for a Multiplex Photoelectrochemical Bioassay of Antibiotic-Resistant Genes

“…Currently, novel PCR-free approaches for ARG detection have been developed, though scarce, such as optical sensor, − colorimetry, and electrochemical/photoelectrochemical (PEC) sensor. − Among these methods, the PEC assay has gained intensive attention, profiting from the different energy forms of the excitation and the detection signals with the merits of remarkable sensitivity, and inherent miniaturization. − The inherent maneuverability of the PEC assay, for example, the engineering of the photoactive materials, the flexibility of the applied excitation wavelength, and potential endows it with great promise as the multiplex assay. And thereafter the light/potentiometric-addressable or wavelength-resolved PEC multiplex assays have been established for multitarget analysis. − However, the processes for partitioning electrodes or fabricating electrode arrays on each tiny area are laborious and impair the sensitivity aroused from the insufficient utilization of light irradiation, limiting the development of multiplex PEC assays.…”

New Trends in Nanoarchitectured SERS Substrates: Nanospaces, 2D Materials, and Organic Heterostructures

“…[111,133] Diazonium modification provides unique opportunities for the postfunctionalization of organic functional groups and the design of materials with high selectivity. The availability of terminated DNA, [134] antibodies, [135] and polymers [136] provides easy cross-coupling functionalization protocols for selective and sensitive SERS sensing. Simultaneously, the growing trend of surface-assisted growth of nanoporous layers requires the preliminary grafting of anchoring groups like COOH [137] or CN, [138] which are the most common building blocks for MOFs/COFs.…”

Raman spectrum matching with contrastive representation learning

“…Deep neural networks have achieved superior performance across many fields. 2,[16][17][18][19][20][21][22] In the context of spectral matching, neural networks are usually trained in a supervised manner, and optimized directly to map an observed spectrum to a score for each reference substance in the training database. 4,5,7,23 Typically a convolutional neural network (CNN) is used since it can extract the fingerprint characteristics better, leading to a higher analyte identification accuracy.…”