Skip to content

BRENDA 2026.1 (build 1)

Choose a tag to compare

@edkerk edkerk released this 17 Oct 22:25
· 3 commits to main since this release
b864c97

BRENDA files for GECKO and geckopy, from BRENDA release 2026.1 (CC BY 4.0; Chang et al., BRENDA, https://doi.org/10.1093/nar/gkaa1025). Changes from the BRENDA data: values are converted to kcat in 1/s, grouped (kcat, sa, mw) or screened for unit errors (measurements).

File Contents Used by
kcat.tsv.gz kcat per EC number, substrate and organism: maximum, median, count, references fuzzy matching (kcatsFromBrenda)
sa.tsv.gz specific activity per EC number, substrate and organism fuzzy matching
mw.tsv.gz molecular weight per EC number and organism fuzzy matching
measurements.tsv.gz 57,037 wild-type kcat measurements with an EC number, one row each, after the outlier screen; with substrate InChIKey, organism, UniProt accession, sequence id, temperature, pH, PubMed id EC tree (kcatsFromECTree)
sequences.fasta.gz the 6,577 sequences of the measurements, by sequence id homologs
substrates.tsv.gz ChEBI, MetaNetX and name keys per substrate InChIKey matching model metabolites
kcatCombineParams.tsv bias and error variance per source (EC tree, homologs, OpenKineticsPredictor methods) and evidence bin, fitted on the kcat benchmark's development split combine stage

The outlier screen leaves out values more than four orders of magnitude below their EC class, and clusters of values below 0.01 /s that sit three orders of magnitude under the other values of their EC number (423 values).