Novel Approach to Freshwater Diatom Profiling and Identification Using Raman Spectroscopy and Chemometric Analysis

(1) An approach with great potential for fast and cost-effective profiling identification of diatoms in lake ecosystems is presented herein. This takes advantage Raman spectroscopy. (2) The study was based on the analysis 790 spectra from 29 species, belonging to 15 genera, 12 families, 9 orders 4 subclasses, which were analysed using chemometric methods. data first by a partial least squares regression discriminant (PLS-DA) characterise diatom species. Furthermore, method developed streamline integrated interpretation PLS-DA when high number significant components extracted. Subsequently, an artificial neural network (ANN) used taxa data. (3) PLS produced profile each species reflecting its biochemical composition. ANN models useful identify various accuracy. (4) Compared studies literature, involving huge datasets one four magnitude larger than ours, sensitivity found Achnanthidium exiguum (67%), Fragilaria pararumpens Amphora pediculus (71%), minutissimum (80%) Melosira varians (82%).