**Preface to "Metabolomics Data Processing and Data Analysis—Current Best Practices"**

Metabolomics analysis has taken its place as a staple tool in all research areas across the bioscience and medical scientific fields where chemical matter is involved. While technical developments on the instruments used for metabolomics analytics allow for deeper than ever chemical exploration of biological samples, the importance of appropriate data-analytical approaches to treat, analyze, and interpret the vast metabolomics data is increasingly highlighted. The data-analytical workflow required for metabolomics study is a multi-step procedure necessitating different software and algorithm approaches for different steps that include but are not limited to peak picking, data preprocessing, and metabolite annotation and identification as well as visualization. In this book, we present a collection of papers focusing on practices and resources for various aspects of the metabolomics data-analytical workflows, starting from data collection all the way to the presentation of publication-ready metabolomics results, including both reviews on the current best practices as well as reports describing novel, innovative approaches for aspects such as in silico prediction of metabolite structures. Any metabolomics study is a multidisciplinary effort necessitating expertise across areas of, e.g., analytical chemistry, biochemistry, bioinformatics, and data-analytics. Therefore, fluent combination of the various steps involved in the workflow involves various challenges and factors to be taken into consideration. Here, current practices are reviewed, from samples to biochemical interpretation (Ivanisevic and Want, 2019), advances in metabolic modeling on a genomic scale (Sen and Oresiˇ c, 2019), as well as possibilities for open-source ˇ algorithms for data-analysis (Klavus et al. 2020). In addition, is the presentation of a web-based ˚ interface that fosters many parts of the metabolomics statistical workflow (Chong et al. 2019). As described by Ivanisevic and Want, metabolite identification remains as one of the pitfalls of metabolomics analysis, and it is therefore essential that advanced procedures are developed for both data acquisition (DIA and DDA) as well as MSMS annotation and metabolite identification. In this book, novel strategies for the computational prediction of mass fragmentation spectra (Djoumbou-Feunang et al. 2019), the integration of computational predictions provided by various algorithms to foster the in silico prediction of metabolite structures (Ernst et al. 2019), handling of DIA MS/MS spectra (Peris-D´ıaz et al. 2019), as well as an in silico framework to optimize the acquisition of mass fragmentation data (Wandy et al. 2019) are presented. Databases and data repositories are an inevitable part of an efficient metabolomics analysis workflow. Curated databases of spectral libraries encompassing both experimental and in silico predicted fragmentation spectra are a prerequisite for efficient metabolite identification. This book contains a section where databases were reviewed and gaps in coverage in terms of different metabolite classes were identified (Oberacher et al. 2018; Frainay et al. 2018). Repositories holding raw metabolomics data enable further utilization of data collections and enable efficient collaborative attempts as described in Tsugawa et al. 2019. Within metabolomics data-analysis an essential element is the efficient strategy for feature selection benefitting especially from the multivariate nature of metabolomics data. Therefore, both advanced methods for the chemometric processing of the data, as well as visualization of the results with ease of interpretation of the biological significance are focus areas requiring further development, as described in the final three papers of this book (Schillemans et al. 2019; Gao et al. 2019; del Castillo et al. 2019). We hope the book will serve as useful resource for anyone entering the field of metabolomics and, especially, the data-analytical part of the technology. Likewise, the book presents various novel algorithms and combined pipelines that may well be utilized by experienced researchers in the field, as well as be further developed owing to the open-source nature of all the presented resources.

#### **Justin J.J. van der Hooft, Kati Hanhineva**

*Editors*

*Review*
