MetabDBHuman Metabolism Database

Studies

Loading studies…

How data are connected in MetabDB

①–⑥ are stored link types. Evidence layers describe the analyses supporting each link.

How data are connected in MetabDBFive entity groups E1 to E5. Stored links 1 to 4 connect variants or genomic contexts with metabolic measures. Link 5 connects genes, and link 6 connects diseases or GWAS traits, to metabolic measures. The dashed protein connection is supporting pQTL evidence, not a seventh stored relationship. Connections have no causal arrows.①②③④⑤⑥Association: incident · prevalent · mortalityGenetic evidence: colocalisation · MRpQTL colocalisationE1Variants / genomic contextsVariants · groups · evidence setsE2GenesRare variant / burden · gene-linked QTLE3Protein targetsProtein identity · pQTL evidenceE4Metabolic measuresMetabolic measures · lipids · ratiosE5Diseases / GWAS traitsOutcomes · other GWAS traits
Lines show data links, without a causal direction.

Entities · E1–E5

E1 Variants / genomic contexts

E2 Genes

E3 Protein targets

E4 Metabolic measures

E5 Diseases / GWAS traits

Stored links · ①–⑥

①Variant–measure
②Representative variant–measure
③Position group–measure
④Evidence set–measure
⑤Gene–measure
⑥Disease / trait–measure

⑤ and the dashed pQTL link belong to the gene/protein evidence group. The dashed link is supporting evidence.

Evidence relationships

Entity pairEvidence
Variant / genomic context ↔ metabolic measureGWAS / metQTL · fine-mapping
Gene ↔ metabolic measureRare variant / burden · eQTL colocalisation
Protein ↔ metabolic measurepQTL colocalisation
Metabolic measure ↔ disease / outcomeDirect associationIncident · Prevalent · Mortality
Metabolic measure ↔ disease / GWAS traitGenetic evidenceColocalisation · MR

①–④ organize genomic records at different levels. Endpoint contexts apply to direct associations; evidence layers can support the same stored link.

Study source catalog

Study order: NMR, MS, Lipidomics; within each assay, non-UKB primary cohorts before UKB, then first online publication date. Secondary UKB use is retained below.

Measures = retained source measurement IDs; Compounds = compound-labelled subset (lipids classified separately). UK Biobank † = validation or follow-up use.

Measurement platforms

Assay / platformStudiesWhat was measured
Nightingale NMRS1–S11Plasma/serum metabolic measures, including lipoproteins, concentrations and ratios
MSS12–S15Broader small-molecule measurements in CSF, plasma, serum or urine
LipidomicsS16Detailed plasma lipid species

NMR studies contribute most of the genetic relationship records, whereas MS and lipidomics studies add broader chemical and sample coverage.

Clinical endpoints and evidence

Clinical endpoints and evidence by source study.

① Variant–measure② Representative variant–measure③ Position group–measure④ Evidence set–measure⑤ Gene–measure⑥ Disease / trait–measure

Colocalisation = colocalisation; MR = Mendelian randomisation. * Rare variant / burden marks published analyses; other evidence columns mark retained data. S3 endpoints use the full UKB reference panel.

Study statistics

Explore entity coverage, all six relationship layers, source results and study overlap in Summary → Studies.

Study overlap

Compare measurement coverage and shared evidence between studies. Select a cell for details.

Overlap uses standardized measurement IDs or entity pairs. Source-only and unresolved records remain in downloads but are excluded from canonical overlap. Study overlap does not imply independent replication.

How are pairs counted?