Skip to content

Metabolomics 101 (cancer)

Under scientific review

This introductory page is under scientific review. Use it for orientation, not clinical decision-making.

Metabolomics profiles small molecules and flux through metabolic networks. Tumors rewires metabolism (Warburg effect is the classic example; modern work is far richer).

The Challenge of Chemical Heterogeneity

Unlike DNA, which is composed of just 4 predictable bases, metabolites are incredibly diverse molecules. They vary drastically in:

  • Size: From small ions to long lipids.
  • Polarity: Water-loving (hydrophilic) molecules vs water-hating (hydrophobic) molecules.
  • Charge: Acids, bases, neutrals.

Computational Consequence: There is no single machine capable of reading the entire "metabolome". Data engineers often need to fuse data coming from different hardware platforms:

  • LC-MS (Liquid Chromatography-Mass Spectrometry): Excellent for large and polar metabolites.
  • GC-MS (Gas Chromatography-Mass Spectrometry): Used for volatile molecules (e.g., short-chain fatty acids).
  • NMR (Nuclear Magnetic Resonance): Less sensitive than MS, but does not destroy the sample and is highly reproducible quantitatively.

The Annotation Problem (Identification)

In genomics, you align a read against the Reference Genome. In metabolomics, alignment is done against chemical databases, which is notoriously difficult:

  • The machine detects an "Exact Mass" of $180.0634$ Da.
  • That could be Glucose, but it could also be Fructose, or Galactose. They have the exact same chemical mass (Isomers).
  • The Bottleneck: Data scientists need to use not just the mass, but the "Retention Time" (the time the molecule took to exit the LC-MS column) and databases like the HMDB (Human Metabolome Database) or METLIN to guess the identity. Often, 70% of the signals in a metabolomic matrix remain as Unidentified Features.

Fluxomics: Adding the Fourth Dimension (Time)

Measuring the static level of Glucose in a tumor does not tell if it is sitting there or being furiously consumed. To solve this, we use computation coupled with Stable Isotopes:

  1. The patient or cell receives a "heavy" version of the nutrient (e.g., Glucose with Carbon-13 instead of normal Carbon-12).
  2. Cancer cells devour this 13C-Glucose and break it down into Lactate, Glutamate, etc.
  3. The Mass Spectrometer can read the weight difference (1 extra unit per carbon).
  4. Isotope Tracing Algorithms: Software mathematically tracks how the fractions of heavy carbon flowed through the Cancer Metabolic Network over time. This reveals the true "bottlenecks" of the tumor engine, indicating the best targets for new metabolic drugs.

See also

Early public release. Content evolves through continuous review. Questions: [email protected] · CC BY 4.0 where applicable.