Frequently Asked Questions
The quality of analytical data is the key to make the right decisions at the right time. As an analytical chemist doing regular LC/ESI/MS analysis, you have probably found yourself in a situation where it is nearly impossible to quantify the analyte due to a lack of analytical standards. Quantem provides the possibility to quantify compounds for which analytical standards are not available in the lab; therefore, opening up new possibilities in non-targeted screening, suspect screening, and other applications where analytical standards are not available. The Quantem approach, based on machine learning and years of mass spectrometry research by the core team members, will allow you to make better decisions on prioritization in metabolomics, drug development as well as in environmental applications. Besides, Quantem can make your data quantitative retrospectively: it can be applied to your already collected LC/MS spectra.
For years, researchers and analysts have relied on targeted analysis methods. In spite of high accuracy of the obtained results, this has lead to overlooking a number of pharmaceutical byproducts, important metabolites, and environmental contaminants. The suspect and non-targeted analysis with LC/HRMS have come to change this. The non-targeted analysis allows detecting hundreds or thousands of compounds without preselection of the analytes.
The most significant obstacles for non-targeted LC/HRMS screening today are the scarcely defined scope of the method and, more importantly, the inability to provide quantitative information. This is well shown by the data. Currently, the Human Metabolome Database contains 114 100 compounds and only 3383 (<3%) of these have been both detected and quantified. Due to the intensive developments in the field of compound identification tools, the proportion of compounds that can be identified with non-targeted methods is constantly increasing. The inability to quantify these compounds is a problem not only in metabolomics, but also in pharmaceutical analysis, environmental analysis, and even in monitoring illegal substances.
The inability to provide quantitative data for non-targeted screening with LC/ESI/HRMS originates from the vastly different ionization efficiencies of different compounds in ESI source. Overall, the ionization efficiencies of the compounds detectable with LC/ESI/MS vary up to 100 million times. For example, even relatively similar compounds like the positional isomers 2- and 4-nitrophenol yield a 40 times different response at equal concentrations. Additionally, LC conditions also affect the ionization efficiencies.
If no commercial analytical standards are available, the only choices so far have been to either (1) synthesise the standards in-house which is very expensive and time-consuming or (2) use other compounds for quantification and ignore the possibility of vastly different response factors. The latter choice could lead to errors up to 10 million times. Quantem provides a third, overwhelmingly faster, cost-effective and accurate option.
Quantem Model combines the fundamental research in the field of mass spectrometry with data science to provide the first solution to situations where there simply are no analytical standards available for quantification. Quantem uses machine learning to predict response factors of analytes taking into account the eluent composition at the retention time and instrument you are using as well as the matrix.
With Quantem you can switch to an approach where your quantification is not limited by the availability of analytical standards.
What is more:
- Quantification is fast – quantify around 10 000 peaks in 24h;
- Quantification is cost-effective – you won’t need to buy expensive analytical standards or spend time and money on synthesising them;
- Quantification is accurate – average quantification error is less than 5 times;
- Quantification can be done retrospectively – you can easily quantify your analysis data acquired even years ago;
- Direct comparison between standard-substance-free analysis results obtained on different instruments and even in different labs opening the door for large scale collaboration in the field of quantitative non-target analysis.
- The concentration can be estimated even if you have multiple candidate structures.
- Quantem approach can be implemented immediately without any further modifications to your current analysis practice and is easy to use.
Yes, Quantem model takes into account the variability between instruments making the prediction result specific to every client. This is established since Quantem Model is developed and tested on many different intruments including Agilent, Thermo, Waters, Sciex and Bruker.
Yes, Quantem model takes into account the variability between analysis methodologies making the prediction result specific to your analysis. This is established since Quantem Model is developed and tested under many different conditions.
Quantem Model is applicable for:
- All common eluent compositions, both in terms of organic modifiers and additives;
- Both positive and negative mode ESI;
- Gradient elution, including different flow rates from microflow up to 1mL/min.
Yes, Quantem model takes into account the variability between analytes and samples making the prediction result specific to your analysis. This is established since Quantem Model is developed and tested in different matrices with thousands of diverse analytes.
Quantem Model is applicable for:
- Numerous types of analytes with logP from -10 to +10 and molar mass below 1500 Daltons;
- Different matrices, e.g. biological samples (urine, plasma, etc.), food samples (cereal, etc.), environmental samples, plant-based materials, etc..
For the quantification with Quantem, you will need 5 compounds with known concentration that are measured under the same conditions as the analytes. We recommend to choose 5 or more compounds that are evenly divided over the chromatogram and have as different ionization efficiency (IE) values as possible. The frequency of measuring these compounds (e.g every run/day/week) depends on your application. You can choose and use your in-house compounds, or we can provide a mixture that suits your needs.
Yes and no.
Yes, Quantem model uses SMILES code to predict the concentration, thus you will have to know the structure of the compound.
But, if you have several structure candidates, you can calculate the concentration for all of them without any meaningful time-loss.