Evaluation of Discrimination Performance in Case for Multiple Non-Discriminated Samples: Classification of Honeys by Fluorescent Fingerprinting
| Autor: | Andrei V. Garmash, Ivan N. Krylov, Vladislav S. Orekhov, Evgenii V. Skorobogatov, Elizaveta A. Rukosueva, Valeria Belikova, Mikhail K. Beklemishev |
|---|---|
| Rok vydání: | 2020 |
| Předmět: |
Fluorophore
principal component analysis honey lcsh:Chemical technology 01 natural sciences Biochemistry Article Analytical Chemistry Chemometrics chemistry.chemical_compound Sensor array fluorophores lcsh:TP1-1185 Electrical and Electronic Engineering fluorescent fingerprinting Instrumentation Mathematics Mahalanobis distance 010405 organic chemistry 010401 analytical chemistry Thiazole orange chemometrics Fluorescence Fluorescence spectra Atomic and Molecular Physics and Optics 0104 chemical sciences chemistry tris(2 2′-bipyridyl)dichlororuthenium(II) Principal component analysis Biological system discrimination |
| Zdroj: | Sensors (Basel, Switzerland) Sensors Volume 20 Issue 18 Sensors, Vol 20, Iss 5351, p 5351 (2020) |
| ISSN: | 1424-8220 |
| Popis: | In this study we develop a variant of fluorescent sensor array technique based on addition of fluorophores to samples. A correct choice of fluorophores is critical for the successful application of the technique, which calls for the necessity of comparing different discrimination protocols. We used 36 honey samples from different sources to which various fluorophores were added (tris-(2,2&prime bipyridyl) dichlororuthenium(II) (Ru(bpy)32+), zinc(II) 8-hydroxyquinoline-5-sulfonate (8-Ox-Zn), and thiazole orange in the presence of two types of deoxyribonucleic acid). The fluorescence spectra were obtained within 400&ndash 600 nm and treated by principal component analysis (PCA). No fluorophore allowed for the discrimination of all samples. To evaluate the discrimination performance of fluorophores, we introduced crossing number (CrN) calculated as the number of mutual intersections of confidence ellipses in the PCA scores plots, and relative position (RP) characterized by the pairwise mutual location of group centers and their most distant points. CrN and RP parameters correlated with each other, with total sensitivity (TS) calculated by Mahalanobis distances, and with the overall rating based on all metrics, with coefficients of correlation over 0.7. Most of the considered parameters gave the first place in the discrimination performance to Ru(bpy)32+ fluorophore. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|