Login / Signup

Total protein assay by PCA-based RGB-spectrum conversion methods with smartphone-acquired digital images.

Arinori InagawaMiwa KimuraNobuo Uehara
Published in: Analytical sciences : the international journal of the Japan Society for Analytical Chemistry (2022)
Total protein concentrations in the aqueous solutions were determined from the absorption spectra reproduced from smartphone-captured digital color images. We employed two different procedures for protein determination: the pyrogallol red molybdate method and Bradford's method. The principal-component-analysis-based reproduction process, which was previously reported by our research group, enabled the conversion of RGB values to score values for a linear combination of loading vectors to generate reproduced absorption spectra. The reproduced spectra were identical to those measured using a commercially available spectrophotometer. The total protein assays of commercial soymilk and human serum samples were carried out with both coloration reagents, and the obtained results were in good agreement with those attained using a conventional spectrophotometer. These results show that the proposed method enables smartphone-based ratiometric analysis of real samples without requiring any monochromating equipment.
Keyphrases
  • protein protein
  • deep learning
  • amino acid
  • binding protein
  • high throughput
  • optical coherence tomography
  • quantum dots
  • machine learning