Special Issue "Signal Processing and Machine Learning for Metabolomics"

A special issue of Metabolites (ISSN 2218-1989).

Deadline for manuscript submissions: 31 July 2019

Special Issue Editors

Guest Editor
Dr. Santiago Marco

Department of Electronic and Biomedical Engineering, University of Barcelona, Martí i Franquès 1, 08028 Barcelona, Spain;
Signal and Information Processing for Sensing Systems, Institute for Bioengineering of Catalonia, Baldiri Reixac 4-8, 08028 Barcelona, Spain
Website | E-Mail
Interests: chemical sensors; signal processing; machine learning; chemometrics; microsystems; ion mobility spectrometry; metabolomics
Guest Editor
Dr. Jesús Brezmes

Spanish Biomedical Research Centre in Diabetes and Associated Metabolic Disorders (CIBERDEM), Madrid, Spain
Website | E-Mail
Interests: signal processing; data analysis; clustering; pattern recognition; artificial neural networks; deep learning; metabolomics; pre-processing; GC-MS and GCxGC-MS; Maldi-TOF
Guest Editor
Dr. Alexandre Perera Lluna

1 Automatic Control Department, Universitat Politècnica de Catalunya, C/Pau Gargallo 5, 08028 Barcelona, Spain
2 Centre de Recerca en Enginyeria Biomèdica, Universitat Politècnica de Catalunya, Pau Gargallo 5, Barcelona, 08028, Spain
3 CIBER of Bioengineering, Biomaterials and Nanomedicine (CIBER-BBN), Barcelona, Catalonia, Spain
Website | E-Mail
Interests: bioinformatics; bioengineering; signal processing; machine learning; pattern recognition

Special Issue Information

Dear Colleagues,

Metabolomics has witnessed astonishing advances in the last decade, based in new instrumental developments in mass spectrometry and magnetic nucelar resonance, with profound implications in life sciences. These instruments provide rich and complex signals, where the relevant information is hidden among an incredible amount of noise related to other sources of variance (lifestyle, diet, genomics variablity) different from the source of interest (related to health condition, therary or toxic exposure). The analysis of these signals is particularly challenging in untargeted metabolomics, aiming to make an exhaustive analysis of the complete metabolome available in a certain body fluid. Metabolite identification with libraries and databases is hindered by noise both in mass spectrometry and nuclear magnetic resonance.

While the analysis of serum and urine is the main target, metabolomics is expanding to the analysis of other fluids such as breath, tears, sweat, faeces, intestinal gases, vaginal discharges and others. In these cases, signal analysis and interpretation is hindered by incomplete understanding of the metabolome, and additional sources of variability related to the absence of standarized methods for sample collection and storage prior to analysis.

Beyond the traditional mature analytical techniques many emerging instrumental techniques are also entering the metabolomics field. Examples of these technologies are chemical sensors, but also miniature spectrometers based on Ion Mobility, direct injection mass spectrometry, and spectroscopies, such as NIR, Visible or Raman. Recent technological advances allow that the principles of measurement, only available in desktops, in analytical lab instrumentation, are now handheld instruments, operating in point-of-care scenarios, with on-board signal and data processing. This novel instrumentation provides signals or spectral time-series that require online signal and data processing in order to obtain the final desired results.

A variety of signal processing problems appear in this domain and they require different signal and data processing tools. Most spectral signals require baseline correction, noise removal, peak alignment, peak dectection and peak deconvolution. Here there is a lot of interest in matrix factorization techniques such as NMF but also higher order tensorial methods, such as PARAFAC, Tucker 3, and others, with a variety of constraints related to prior information on the solution. Beyond that, systems require solutions for drift compensation, calibration transfer, or chemical noise (interferants) rejection. In terms of self-diagnostics, we need techniques for fault detection, identification, and diagnosis. After preprocessing, data is further analyzed using machine learning techniques mostly for classification but also regression problems. In addition, optimization techniques, required for optimal feature extraction or selection (biomarker discovery), are also needed during algorithmic development. The development of methodologies for system validation to ensure proper generalization to new samples is equally important. Last, but not least, the use of design of experiment techniques in system calibration or recalibration also require further research.

We solicit contributions focused on signal processing and machine learning for metabolomics in the following, or similar, areas:

  • Signal pre-processing: Digital filters, baseline corrections, peak detection and alignment
  • Peak deconvolution using matrix factorization techniques: NMF, MCR-ALS, Hard and soft-constraints
  • Tensorial decomposition methods for N-way data: PARAFAC, PARAFAC2, Tucker3
  • Orthogonal Projection Filters: Component Correction, Net Analyte Signal, Orthogonal Signal Correction,  OPLS, O2PLS
  • Fault detection, identification and diagnosis
  • Counteraction of instrumental shifts
  • Feature Extraction and Selection
  • Classification and Regression techniques
  • Validation techniques
  • Novelty detection
  • Calibration transfer
  • Development of applications with novel approaches for signal processing and machine learning
  • Applications of metabolomics strategies in related areas such as Foodomics.
  • Development of Software Packages in R, Python, MATLAB for signal processing and machine learning

Dr. Santiago Marco
Dr. Jesús Brezmes
Dr. Alexandre Perera Lluna
Guest Editors

Manuscript Submission Information

Manuscripts should be submitted online at www.mdpi.com by registering and logging in to this website. Once you are registered, click here to go to the submission form. Manuscripts can be submitted until the deadline. All papers will be peer-reviewed. Accepted papers will be published continuously in the journal (as soon as accepted) and will be listed together on the special issue website. Research articles, review articles as well as short communications are invited. For planned papers, a title and short abstract (about 100 words) can be sent to the Editorial Office for announcement on this website.

Submitted manuscripts should not have been published previously, nor be under consideration for publication elsewhere (except conference proceedings papers). All manuscripts are thoroughly refereed through a single-blind peer-review process. A guide for authors and other relevant information for submission of manuscripts is available on the Instructions for Authors page. Metabolites is an international peer-reviewed open access monthly journal published by MDPI.

Please visit the Instructions for Authors page before submitting a manuscript. The Article Processing Charge (APC) for publication in this open access journal is 1000 CHF (Swiss Francs). Submitted papers should be well formatted and use good English. Authors may use MDPI's English editing service prior to publication or during author revisions.


  • Baseline Corrections
  • Instrumental shift compensation
  • Peak deconvolution
  • Peak detection
  • Metabolite identification

Published Papers (1 paper)

View options order results:
result details:
Displaying articles 1-1
Export citation of selected articles as:


Open AccessArticle
Using Expert Driven Machine Learning to Enhance Dynamic Metabolomics Data Analysis
Metabolites 2019, 9(3), 54; https://doi.org/10.3390/metabo9030054
Received: 18 January 2019 / Revised: 5 March 2019 / Accepted: 18 March 2019 / Published: 20 March 2019
PDF Full-text (2744 KB) | HTML Full-text | XML Full-text | Supplementary Files
Data analysis for metabolomics is undergoing rapid progress thanks to the proliferation of novel tools and the standardization of existing workflows. As untargeted metabolomics datasets and experiments continue to increase in size and complexity, standardized workflows are often not sufficiently sophisticated. In addition, [...] Read more.
Data analysis for metabolomics is undergoing rapid progress thanks to the proliferation of novel tools and the standardization of existing workflows. As untargeted metabolomics datasets and experiments continue to increase in size and complexity, standardized workflows are often not sufficiently sophisticated. In addition, the ground truth for untargeted metabolomics experiments is intrinsically unknown and the performance of tools is difficult to evaluate. Here, the problem of dynamic multi-class metabolomics experiments was investigated using a simulated dataset with a known ground truth. This simulated dataset was used to evaluate the performance of tinderesting, a new and intuitive tool based on gathering expert knowledge to be used in machine learning. The results were compared to EDGE, a statistical method for time series data. This paper presents three novel outcomes. The first is a way to simulate dynamic metabolomics data with a known ground truth based on ordinary differential equations. This method is made available through the MetaboLouise R package. Second, the EDGE tool, originally developed for genomics data analysis, is highly performant in analyzing dynamic case vs. control metabolomics data. Third, the tinderesting method is introduced to analyse more complex dynamic metabolomics experiments. This tool consists of a Shiny app for collecting expert knowledge, which in turn is used to train a machine learning model to emulate the decision process of the expert. This approach does not replace traditional data analysis workflows for metabolomics, but can provide additional information, improved performance or easier interpretation of results. The advantage is that the tool is agnostic to the complexity of the experiment, and thus is easier to use in advanced setups. All code for the presented analysis, MetaboLouise and tinderesting are freely available. Full article
(This article belongs to the Special Issue Signal Processing and Machine Learning for Metabolomics)

Figure 1

Metabolites EISSN 2218-1989 Published by MDPI AG, Basel, Switzerland RSS E-Mail Table of Contents Alert
Back to Top