Professor Roy Goodacre FRSC, FLSW, FSAS

Metabolomics involves the study of the biochemical processes that involve metabolites. These small molecules are chemically diverse and thus require detailed bioanalytical measurements. The research in our group is aimed at developing robust and reproducible metabolomics platforms that can be used in clinical and plant studies, as well as for understanding microbial systems.

We have generated a set of standard operating procedures (SOPs) for gas chromatography-mass spectrometry (GC-MS) and liquid chromatography-mass spectrometry (LC-MS). These SOPs involve the novel use of quality control (QC) standards that can be used to assess unavoidable instrument drift and to correct for it. The use of QCs allows us to combine metabolomics data over many months to years, and we used these to profile serum from large populations in the UK.

These SOPs are available in Nature Protocols and a summary of these and other protocols are found within our BioSpec resources. A recent review from 2018 is available in Metabolomics.

We use metabolomics in many different areas and this include the following studies:
- Human and mammalian biology and diagnostics
- Microbial spoilage and pathogen contamination of food
- Synthetic biology applications focusing on recombinant protein production
- Bacterial exposure to human drugs and antibiotics as well as AMR
- Plant systems for enhanced carbon fixation
- Volatile profiling for breathomics

Papers highlighting these studies are found under the publications tab.

Raman spectroscopy measures the vibrations of bonds within functional groups, and usually involves the exchange of energy with a monochromatic light source such as a laser. This results in the construction of a Raman ‘fingerprint’ of the sample which can be used to identify and quantify biochemical components within mixtures. Sample analysis can be performed on bulk material, through containers as well as via microscopy for chemical imaging, with resolution in the micrometre range.

In 2004 we published the first paper that demonstrated that surface enhanced Raman scattering (SERS) could be used for the discrimination of bacteria. SERS is a method that enhances the usually weak Raman effect and we have been concentrating on using design of experiments (DoE) based on fractional factorial design (FFD) combined with multi-objective evolutionary algorithms for SERS optimisation. SERS has then been used in a variety of different areas for quantitative analysis, and for a review on the topic for absolute quantification was published this year in TrAC.

In addition to SERS we have investigating other enhancement techniques based on deep UV resonance Raman for synthetic biology. We have also very recently been investigating spatially offset Raman spectroscopy (SORS), as this is able to take measurements through barriers for detecting counterfeit alcohol without opening the (e.g.) whisky bottle.

We have been developing a range of diverse Raman spectroscopy and related techniques in many different areas:
- Single bacteria are being analysed for identification as well as functional analysis via the adoption of stable isotope labeling
- SERS has been developed for trace quantitative analysis of drugs and metabolite
- LC-Raman/SERS is being developed for metabolomics
- Raman microscopy allows chemical mapping for imaging the location of drugs in eukaryotic cells
- Deep UV Resonance Raman at 244 nm for real time on-line whole cell biotransformations for synthetic biology
- Portable Raman and SORS instruments are being deployed for food security

Papers highlighting these studies are found under the publications tab.

Metabolomics and Raman spectroscopy, and indeed all omics, generate large amount of data and we have been developing a range of methods for data preprocessing and pretreatment prior to multivariate and univariate analyses.

Methods we have used and developed include:
- Traditional methods like PCA, HCA, DFA (LDA/CVA), PLSR and PLSDA.
- Multiblock approaches for data containing multi influential factors.
- Procrustes analysis.
- Design of Experiments (DoE) including fractional factorial analysis, genetic algorithms, Pareto optimality, for MS and Raman/SERS.
- We’re also very keen on validation as this is essential and we achieve this via resampling: k-fold cross validation, bootstrapping incorporating permutation testing.

We try and help the community by being open access:
- Our Matlab and R code freely available via GitHub
- A popular tutorial on life beyond PLS-DA for metabolomics
- Metabolomics data standards for data sharing and data analysis, with plans to effect the same in infrared and Raman.