作者: Kirill A. Veselkov , James S. McKenzie , Jeremy K. Nicholson
DOI: 10.1002/9780470034590.EMRSTM1407
关键词:
摘要: High-resolution NMR spectroscopy is applied for molecular phenotyping across a range of pharmaceutical and clinical applications such as drug toxicity, disease diagnostics, personalized healthcare studies. A typical profile biological sample contains tens thousands signals arising from hundreds endogenous exogenous metabolites. The generated data requires advanced computational workflows to translate raw spectroscopic into pharmacology clinically useful information. This article outlines various chemoinformatics strategies that maximize pharmacologically relevant information recovery one-dimensional spectra samples. In broad terms, the outlined involve (i) analytical signal preprocessing improved recovery, (ii) multivariate statistical explorative predictive analyses spectra, (iii) time-course address pharmaceutically questions. Keywords: chemoinformatics; NMR spectroscopy; multivariate analysis; signal processing; timecourse analysis