Glossary
Spectral deconvolution
Real samples are crowded. When compounds elute together, their signals overlap, and deconvolution is what assigns each fragment and ion to the compound it actually came from.
Definition
Deconvolution is the process of separating overlapping signals in LC/MS data so that each is attributed to the correct compound. When two or more analytes elute at nearly the same time, their ions and fragments mix in the recorded spectra, and grouping them by shared elution profile is what reconstructs a clean spectrum for each compound. Deconvolution is closely tied to feature grouping, because a real compound produces a set of correlated ions, isotopes, and adducts that rise and fall together. Getting it right reduces false features and improves the quality of any annotation attempted afterward.
Worked example
In practice
Two coeluting metabolites can produce a mixed spectrum, and deconvolution uses the shared rise and fall of each compound's ions over time to split that spectrum into two interpretable ones.
What matters
Where this makes a difference
Coelution is the norm
Complex matrices rarely resolve every compound chromatographically, so overlapping signals are expected. Deconvolution is how mixed spectra become compound level information.
Grouping correlated ions
A compound produces isotopes, adducts, and fragments that share an elution profile. Grouping by that shared profile reduces the feature count to something closer to real compounds.
Effect on annotation
Clean, deconvolved spectra give annotation tools a better target than a mixed spectrum, so deconvolution quality propagates into identification confidence.
Keep reading
Related
Want to see the method on real data?
Read what a dataset assessment involves and what you get back.
Grounded in the same first principles that Metablify is built on.