Comparison of subcellular responses for the evaluation and prediction of the chemotherapeutic response to cisplatin in lung adenocarcinoma using Raman spectroscopy

Multivariate statistical methodologies applied in biomedical Raman spectroscopy: assessing the validity of partial least squares regression using simulated model datasets

“…The spectral data (X matrix) is thus related to the targets (Y matrix) according to the linear equation Y = XB +E, where B is a matrix of regression coefficients and E is a matrix of residuals. The PLSR algorithms used in this study have been previously published elsewhere 10,11,18,22 and are based on scripts written in house using Matlab 7.2 (The Mathworks Inc.). The algorthim allows for the construction of a regression model which can be used to predict the outcome in a number of different situations.…”

“…In this work, these experimental spectral datasets are employed to construct semi-realistic simulated data to probe the reliability, sensitivity and quantitative nature of these methods when 10,11 Partial Least Squares Regression PLSR is a multivariate statistical method which aims to establish a model that relates the variations of the spectral data to a series of relevant targets. The spectral data (X matrix) is thus related to the targets (Y matrix) according to the linear equation Y = XB +E, where B is a matrix of regression coefficients and E is a matrix of residuals.…”

“…Cisplatin, at varying concentrations in the range 0.05 µM -50µM, was added to cells and Raman microscopic measurements of cells exposed to each dose, including unexposed control, were acquired at a source wavelength of 785nm for both nuclear 10 and cytoplasmic regions 11 . In both cases, multiple spectra were recorded from a total of 60 cells at each exposure level.…”

“…The numerous studies explore applications such as disease diagnostics [1][2][3][4] , cellular imaging [5][6][7][8] , the study of drug [9][10][11] and nanoparticle interactions [12][13][14] on a cellular and sub-cellular level, to name but a few. In both modalities, the spectrum of tissue or cells contains a wealth of information and represents the combined molecular fingerprints of the ensemble of biomolecules contained in the sample.…”

“…Importantly, there are many variants of the PLSR algorithm and, in some instances, hybrid methods which use a combination of two statistical tools in order to extract relevant chemical information have been employed. Although these methods have been applied to a wide range of studies, the details are beyond the scope of this paper although good examples can be found in literature 10,11,18,[21][22][23][24] The potential of Raman spectroscopic microscopy for initial screening of chemotherapeutic efficacy and mechanism of action has been demonstrated by Nawaz et al 10,11,23 . Taking the interaction of cisplatin with the human lung adenocarcinoma cell line, A549, in vitro, as an example, PLSR of Raman spectroscopic datasets was reported to identify and differentiate the direct effects of cisplatin on the cellular biochemistry as a function of drug concentration (dose) and the resultant toxicological response as measured by the MTT cytotoxicity assay.…”