We've updated our Privacy Policy to make it clearer how we use your personal data.

We use cookies to provide you with a better experience. You can read our Cookie Policy here.


Confident Metabolite Identification for Meaningful Results in Multiomics Analyses

Small molecule closeup
Credit: iStock
Listen with
Register for free to listen to this article
Thank you. Listen to this article using the player above.

Want to listen to this article for FREE?

Complete the form below to unlock access to ALL audio articles.

Read time: 2 minutes

Untargeted metabolomic measurements using mass spectrometry are a powerful tool for novel small molecule discovery. The determination of the collision cross section (CCS) has been shown to increase confidence in the annotation and identification of small molecules in various ways.

Publicly available CCS libraries are far from covering the entire chemical space of small molecules. Technology Networks spoke to Heiko Neuweger, director of Bioinformatics Life Science Mass Spectrometry Software R&D at Bruker, to learn about the importance of target compound annotation and identification, the advantages of integrating CCS information into workflows and the benefits of CCS-Predict Pro.

Kate Robinson (KR): Why is the annotation and identification of target compounds important in metabolomics?

Heiko Neuweger (HN): Metabolomics researchers are all too aware of the complexity of chemical space, particularly in the realm of small molecule metabolites. When profiling intricate samples such as natural products, microbial metabolites or lipids, we encounter vast structural diversity that is often convoluted by isomeric and isobaric species. Whether we are looking for bioactive metabolites for drug discovery or uncovering a metabolic pathway involved in a disease state, confident metabolite identification is imperative for results to be meaningful in discovery, quantitation and multiomics analyses.

KR: How does CCS-Predict Pro predict collision cross sections (CCS) for metabolites?

HN: CCS-Predict Pro is an easy-to-use machine learning (ML) model that is accurately trained and validated on measured timsCCS values from thousands of known reference compounds. There are currently two models, one for lipids and one for small molecule metabolites, which can be applied to both positive and negative mode data. CCS-Predict Pro is built into MetaboScape®, Bruker’s untargeted metabolomics data analysis software.

CCS prediction is automatically enabled for any spectral library (e.g., MetaboBase, NIST) or target list with 2D structural information. This means you can take any set of molecules and, using chemical formulas and either InChI or SMILES, a CCS value is predicted and applied to the entire feature table. If a feature matches the predicted CCS value, it is used in annotation quality scoring to show how close the predicted value is to the experimental value.

KR: How accurate are the CCS predictions made by the algorithm?

HN: CCS-Predict Pro is one of the highest-performing CCS prediction algorithms available today. ML models like DeepCCS and AllCCS2 are recently published comparable models. A major benefit of ML models is that they can be continually refined with the addition of more training data. The latest version of CCS-Predict Pro 2024 has further improved its predictive accuracy below one percent CCS error compared with experimentally measured values.

KR: What are the potential applications of this software?

HN: When applied to untargeted metabolomics or lipidomics workflows, CCS-Predict Pro 2024 is a powerful tool to help validate metabolite identities. It is often the case that metabolites don’t have an experimental CCS value to compare to, for example, novel natural product secondary metabolites, drug metabolites, xenobiotics that have been altered by human or microbial metabolism or lipid species with varying carbon chain lengths and double bonds.

Isomeric species are often difficult to separate by chromatography, and depending on the column, mobile phases or the phase of the moon, chromatographic peaks can shift. Isomers also tend to have very similar MS/MS fragmentation patterns that may lead to incorrect annotations. By measuring CCS and using CCS-Predict Pro, along with any spectral library or target list, we can annotate our metabolites and lipids faster and more confidently.

KR: What are the advantages of integrating CCS information into metabolomics workflows?

HN: Unlike retention time, CCS is an intrinsic physical property of a molecule in the gas phase, and it can be used as an orthogonal measurement for identification in untargeted metabolomics and lipidomics workflows.

Since timsTOF instruments can perform direct measurements of mobility (1/K0) with simple calibration, CCS values are accurate and reproducible. Acquiring information-dense liquid-chromatography mass spectrometry with the additional CCS dimension can provide valuable metrics for annotating the metabolome or lipidome in addition to accurate mass, isotope patterns, MS/MS and retention time.

Data analysis tends to be the most challenging part of any metabolomics study, and we all want results as quickly as possible. CCS-Predict Pro was developed to provide an automatic and highly confident workflow to give additional confidence in metabolite and lipid identities untargeted studies.

Heiko Neuweger was speaking to Kate Robinson, Assistant Editor for Technology Networks.