Having reviewed literature on the use of chemometric approaches in quality assessments of medicines (including herbal medicines), I realised that several approaches are adopted. For example, in preprocessing of the data for further analysis, literature reports of methods like normalisation, peak centering, warping, smoothing among others.

Having in mind that the way you preprocess the data may affect the final outcome of the multivariate analysis, I want to find out if there exist any protocol guiding the adoption of any of these tools. For instance, when analysing chromatographic data from HPLC, you may have to correct baseline, then warp and normalise or something. Also, when dealing with FTIR data, you may have to first correct baseline, normalise and smooth (how do you determine the smoothing points?) among others. Are there specific preprocessing tools for specific datasets (that is from different instruments like FTIR, HPLC, LC-MS, etc) and are there specific procedures for the use of such, so that irrespective of who is conducting such analysis, the outcome may always be reproducible?

Thank you.