Webinar Fast Multiclass Meat Metabolomics: Rapid Evaporative Ionisation Mass Spectrometry Meets OPLS-HDA

Fast Multiclass Meat Metabolomics: Rapid Evaporative Ionisation Mass Spectrometry Meets OPLS-HDA

High-dimensional multiclass data can prove challenging and time-consuming to analyse using conventional data analysis methods. Here we revisit an earlier published dataset with 10 different groups (Zhang et al., 2022) to evaluate a new process developed for SIMCA; OPLS-hierachical discriminant analysis (HDA).

Deciphering what impact farming conditions have on meat is important, especially for understanding what combinations of breed, sex and pasture conditions led to the best meat quality and consumer liking. To study how different breed and feed combinations can impact on consumer liking, meat samples from three farms that encompassed four breeds, six pastures, and sex as factors were collected (n=140). Meat pH, colour and tenderness were measured.

Traditional chromatography-mass spectrometry-based metabolomics can provide important molecular insights into biology, but for solid samples such as meat, require time consuming sample preparation. Here we have used an ambient ionisation mass spectrometry method, rapid evaporative ionisation mass spectrometry (REIMS), to measure a fingerprint of the meat samples, based on a high-resolution mass spectrum. Measurements are done directly on a sample without sample preparation and take <10 seconds per measurement.

Initial data analysis was based on pairwise class comparisons using OPLS-DA and VIP-score based variable selection before model confirmation using PCA. This found clear separation between breeds, pasture types and sex, but required building many individual models and took several hours for the data analysis. Using the new OPLS-HDA procedure (Forsgren et al., 2025), the complete pairwise analysis of the 10 classes took <10 minutes, including a robust training-test set comparison and class prediction. We will compare OPLS-HDA with other advanced methods, including MUVR (Shi et al., 2018), a machine-learning based algorithm for unbiased variable selection.

This seminar will outline our experience with using SIMCA^® for metabolomics data analysis using this dataset as an example, and how it compares to other methods for metabolomics data analysis.

What You Will Learn:

Learn how rapid evaporative ionisation mass spectrometry (REIMS) can acquire metabolomics fingerprinting data without the need for sample preparation
Get an insight into how breed and feed types impact on the meat metabolome
Discover how the novel method of OPLS-HDA simplifies the analysis of multiclass metabolomics problems

References:
Forsgren, E., Björkblom, B., Trygg, J., & Jonsson, P. (2025). OPLS-Based Multiclass Classification and Data-Driven Interclass Relationship Discovery. Journal of Chemical Information and Modeling, 65(4), 1762-1770. https://doi.org/10.1021/acs.jcim.4c01799

Shi, L., Westerhuis, J. A., Rosén, J., Landberg, R., & Brunius, C. (2018). Variable selection and validation in multivariate modelling. Bioinformatics, 35(6), 972-980. https://doi.org/10.1093/bioinformatics/bty710

Zhang, R., Realini, C. E., Middlewood, P., Pavan, E., & Ross, A. B. (2022). Metabolic fingerprinting using Rapid evaporative ionisation mass spectrometry can discriminate meat quality and composition of lambs from different sexes, breeds and forage systems. Food Chemistry, 386, 132758. https://doi.org/https://doi.org/10.1016/j.foodchem.2022.132758

Clear form data

Please Complete the Form

First Name *

Last Name *

Email *

Phone Number *

Company *

Job Title *

Address *

Country *

Postal Code *

I would like to sign up for newsletters from Sartorius (Sartorius AG and its affiliated companies) based of my personal interests.
I can revoke my consent at any time with effect for the future by sending an e-mail to unsubscribe@sartorius.com or by clicking on the "unsubscribe" link in e-mails I have received.

* These fields are required Information on the use of your data can be found in our privacy policy

Complete Form to Watch

Presenters

This webinar is co-presented by Alastair Ross and Olle Hartvigsson.

Alastair Ross is a senior scientist and metabolomics lead at AgResearch, a New Zealand government research organisation focused on supporting the agri-food sector. He has a PhD from the Swedish University of Agricultural Sciences in Uppsala, and has worked in the food industry as a scientist in Switzerland, been an academic scientist at Chalmers University of Technology in Sweden, before returning to his home country. He has published over 100 papers and has an H-index of 42. Current research interests include the use of ambient ionisation mass spectrometry to bridge what metabolomics delivers for fundamental science and non-traditional users of metabolomics.

Olle Hartvigsson is a data scientist at AgResearch. He acquired his PhD in computational metabolomics in 2023 from Chalmers University of Technology before making the move to New Zealand. During his PhD he developed a keen interest in multivariate statistics, specifically for its uses in the field of metabolomics, where biological interpretation is frequently the most important outcome. Current research focuses on nitrogen leaching from soils and farmland, life cycle analysis and metabolomics data processing.