Metabolomics data analysis – processing and analysis of a dataset

摘要: Differences in sample quantities, as well experimental variations, require the application of an optimal normalization procedure. Variations metabolite profiles caused by factors beyond investigator’s control and interest may also affect data, lead to erroneous incorrect conclusions. Corrections for these must be made, addition spectral preprocessing and/or quantification. Subsequently, it is possible perform different types analysis, ranging from unsupervised analyses (e.g. determine sub-types or grouping) supervised aimed at classification feature selection. Likewise, quantified metabolic data can integrated with results other experiments, e.g. transcriptomics proteomics, leading development systems biology models. Data analysis strategy depends on specific study there no approach that fit all questions. This chapter provides overview methods used metabolomics their positive negative sides many examples.