Alignment of 1D NMR Data using the iCoshift Tool: A Tutorial
It is well known that several physical factors can affect the absolute and the relative position of NMR signals. Such changes make it difficult to recover useful systematic latent information on large sample sets since, for a given proton population, its chemical shift may vary from sample to sample. Correction of the peak position is therefore a crucial preprocessing step foregoing multivariate data analysis. The interval Correlation Optimized shifting (icoshift) algorithm represents a powerful and versatile tool for dealing with all kinds of signal alignment problems. icoshift allows to choose among a large variety of options, from fully automated corrections of the whole NMR spectrum, to supervised and targeted interventions covering only selected spectral regions. It can automatically cope with intervals of different size and provides automated diagnostics and spectral plots that facilitate easy interpretation of the achieved results. Thanks to its ultra-rapid engine, it is capable of handling huge NMR datasets in reasonable time, allowing working with full spectral resolution and avoiding down-sampling steps, e.g. binning. icoshift is found to be faster and more peak-shape conservative than other recently proposed methods. Furthermore, it has recently been improved to cope also with misalignments in chromatographic datasets, as an advantageous alternative to more traditional and time expensive tools, e.g. COW. This work illustrates how to make the most efficient use of icoshift, using an NMR dataset of foodstuff, showing the different options the tool offers and discussing problems, pitfalls and tricks of the trade in relation to successive multivariate analysis.