Spectralomics – Towards a holistic adaptation of label free spectroscopy

Abstract

Vibrational spectroscopy, largely based on infrared absorption and Raman scattering techniques, is much vaunted as a label free approach, delivering a high content, holistic characterisation of a sample, with demonstrable applications in a broad range of fields, from process analytical technologies and preclinical drug screening, to disease diagnostics, therapeutics, prognostics and personalised medicine. However, in the analysis of such complex systems, a trend has emerged in which spectral analysis is reduced to the identification of individual peaks, based on reference tables of assignments derived from literature, which are then interpreted as biomarkers. More sophisticated analysis attempts to unmix the spectrum of the complex mixture into constituent components, which are then used to characterise the biochemistry of a sample and changes to it, in terms of its constituent components. Data mining the spectra, and in particular change due to kinetic processes, remains a challenge, and it is proposed that the rate of temporal evolution of the combination spectrum can be used in itself as a label by which to guide the spectral analysis. Ultimately, it is argued that the true potential of label free spectroscopy is best harnessed in a truly “spectralomic” approach, by which the spectral signature of an “event”, such as drug intercalation in the DNA of the nucleus of a cell, or a key stage of a cellular pathway such as oxidative stress, is presented. It is envisioned that, in the future, such Spectralomics pathway analysis will be fully integrated with similar omics approaches, potentially ultimately through deep learning algorithms, and underpinned by systems biology kinetic models, to provide a living human cell atlas, describing the function and dysfunction of organism at a cellular level, as the basis for improved healthcare.