Which Sample Matrix Should I Use for My Metabolomics Study?

One of the most common questions asked in metabolomics studies is: “Which sample matrix should I use?” If you’re looking for a simple, clear answer, you probably won’t find it here. However, if you’re interested in understanding why it depends on the situation and how to determine the best starting point for your study, you’ve come to the right place.

Choosing the right sample matrix isn’t about finding the “perfect” option, but about avoiding common pitfalls. Each matrix has its own unique strengths and weaknesses, and choosing the wrong one can severely impact the reliability and interpretability of your data. In this post, we’ll break down the pros and cons of some commonly used matrices and provide practical tips to guide your decision-making process.

Read on to explore how sample matrix selection sets the stage for a successful metabolomics study.

Defining Your Research Goal

Before diving into the details of sample selection, ask yourself the most critical question: "What do you want to achieve with your metabolomics study?" Clearly defining your research objective is the first step to narrowing down your sample matrix options. Once your goal is in focus, the rest often falls into place.  

For instance, if your study involves exploring biological processes that cannot be investigated in humans—such as requiring tissue samples from hard-to-access sites—you’ll likely need to use animal models or alternative systems. On the other hand, if your objective is to identify a stratification biomarker and human biofluids are available, those biofluids naturally become your matrix of choice.  

Similarly, if your study builds on existing data—such as transcriptomics highlighting specific metabolic pathways—you’ll want a matrix and approach that target those pathways. Each research question defines not only the feasibility of metabolomics from a particular matrix but also its relevance to your desired outcome.  

By aligning your choice of sample matrix with your scientific goal, you set a solid foundation for a successful and meaningful study.

Common Sample Matrices in Metabolomics

An overview of advantages and disadvantages:

  Cells and tiss    ues  

Cells and tissues are powerful matrices for studying localized metabolic activity, offering detailed insights into specific cell types or organs. Advances like organoids and organ-on-a-chip systems make this approach even more valuable. However, successful use of these matrices requires careful planning to avoid common pitfalls.

Therefore, there are several key factors to consider:

• While cells and tissues highlight local processes, they may not fully reflect systemic interactions. Findings should be interpreted in the broader context of the organism.  

• Variability within organs can affect results. Combining samples from multiple sites or creating homogenates helps ensure consistent and representative data.  

• Fresh tissue is preferred, as fixation significantly alters metabolite profiles. Immediate processing or proper preservation is crucial.  

• Perfused organs, like the liver, may retain blood, which can skew results. Proper washing or perfusion is needed for tissue-specific profiling.  

• In cell-based studies, culture conditions (media, phenotype changes, co-culturing) can greatly influence metabolic behavior. Optimize these factors for physiologically relevant results.  

• Metabolite diversity requires careful extraction. Single-step methods may work for some studies, but multi-step protocols often yield better results for capturing a range of metabolites.  

By addressing these challenges, cells and tissues can provide precise, actionable insights into specific biological processes.

Blood components

Blood components are among the most widely used matrices in metabolomics, particularly in biomarker discovery and clinical research. As a systemic biofluid, blood reflects metabolic signals from the entire organism, making it invaluable for studying complex diseases and physiological processes.  

Serum and plasma are the most commonly analyzed blood fractions, each with distinct advantages and challenges. Serum is often preferred for its higher metabolite concentrations and suitability for targeted analyses, such as steroid hormone profiling. However, its susceptibility to pre-analytical variations, like oxidation and metabolite release during coagulation, can affect data quality. Plasma, on the other hand, is less prone to these issues, though the choice of anticoagulant (e.g., EDTA) can influence results. Whole blood, while less frequently used due to cellular interference, has niche applications, such as in newborn screening through dried blood spots.  

Innovative sampling methods, such as dried plasma spots, are gaining traction for their logistical advantages, including enhanced stability and ease of transport. Additionally, advanced techniques using sorted blood cells or extracellular vesicles allow researchers to explore specific cellular or molecular processes, such as tumor metabolism or immune regulation.  

Overall, blood components remain a cornerstone of metabolomics research, offering unparalleled access to systemic metabolic insights and expanding possibilities for non-invasive diagnostics and targeted investigations.

Rat serum metabolic profile

Urine

 Urine is a very valuable matrix in metabolomics due to its non-invasive collection and relative stability, even at elevated storage temperatures. These advantages make it an attractive option for large-scale studies and longitudinal monitoring. While 24-hour urine samples are the gold standard for comprehensive analysis, random urine is also common.

The high salt content of urine can pose challenges for analysis, but normalization strategies, such as adjustment for creatinine concentration, can help reduce variability and improve data quality. In diagnostic applications, urine is particularly useful for detecting abnormal metabolites, such as glucose or protein, that can signal potential health issues. However, as a major excretion substance, many metabolites detected in urine are the result of deliberate excretion to maintain systemic homeostasis, which may limit the suitability of the matrix for certain research questions.

Overall, urine is a powerful and accessible metabolomics matrix that provides unique insights into metabolic processes while presenting manageable analytical challenges.

Other biofluids

Other biofluids, such as saliva, tear fluid, cerebrospinal fluid (CSF), bronchoalveolar lavage fluid (BALF), and sweat, offer unique opportunities in metabolomics, though their use is often limited by practical challenges or developmental constraints.  

Saliva and tear fluid are highly accessible due to their non-invasive sampling methods, making them attractive for point-of-care applications. However, metabolomics research in these fluids is still in its early stages. CSF provides a valuable window into central nervous system metabolism, but its invasive collection and ethical constraints on obtaining healthy control samples limit its use. Similarly, BALF is highly relevant for studying respiratory disorders but faces scalability challenges due to the difficulty of sampling. Sweat, while easy to collect non-invasively, remains an emerging matrix with potential yet to be fully realized.  

Despite these hurdles, advances in sampling and analysis are gradually broadening the scope of these biofluids, offering exciting new possibilities for metabolomics research in niche applications.

Other non-invasive sample types

Non-invasive sample types, such as feces, hair, and other emerging matrices like skin lavage, earwax, and nasal mucus, are gaining interest in metabolomics due to their accessibility, ease of collection, and potential for repeated sampling.  

Feces has seen a surge in use, particularly with the rise of microbiome studies. It offers insights into host-microbiome interactions but comes with challenges such as sample variability, dietary confounders, and extraction methods affecting metabolite profiles. While fecal water may focus on cell-free metabolites, harsher extraction methods can capture microbial metabolites, offering a broader coverage. However, the fecal metabolome primarily reflects excreted metabolites, which may limit its ability to represent upper intestinal processes.  

Hair, another non-invasive matrix, provides a long-term record of metabolic changes, capturing weeks or even months of information. While established for measuring certain metabolites like cortisol, its clinical utility in broader metabolomics is less developed due to variability from pigmentation, hair treatments, and the lack of standardized protocols.  

These matrices, while promising, still require further methodological advancements and validation to fully realize their potential in metabolomics research.

Analysis of common metabolites in rats

Conclusion

The question of which sample matrix to use in metabolomics is both essential and complex, as each matrix comes with its own set of advantages and limitations. Whether you choose blood, urine, tissue, or any other biofluid, it’s crucial to align your matrix choice with the specific research objectives and practical considerations of your study.

However, the decision doesn’t have to be an either/or scenario. Combining multiple matrices can provide a more holistic view of the biological process you're studying. For example, integrating plasma and urine analyses in kidney disease research can offer more comprehensive insights than focusing on a single matrix. Similarly, combining blood-based metabolomics with tissue analysis can provide both systemic and localized perspectives, enhancing the overall relevance of the findings.

Even more valuable is the potential for combining metabolomics with other technologies, such as imaging or genomics, to further enrich the data and uncover deeper biological insights.

In the end, the right sample matrix—or combination of matrices—depends on your research question, available resources, and the biological context. Thoughtful selection will lead to stronger, more meaningful results. And while we may not have covered every possible matrix, the exploration of diverse sample types will continue to expand the possibilities of metabolomics research.

Which matrix will you choose for your research? If you have any questions about sample matrix selection, or if you think a matrix we haven't mentioned plays a key role in your work, please contact us.

Reference

1. Shi, Q., Wang, S., Wang, G. et al. Serum metabolomics analysis reveals potential biomarkers of penicillins-induced fatal anaphylactic shock in rats. Sci Rep 14, 23534 (2024). https://doi.org/10.1038/s41598-024-74623-x