Enzyme k_cat, K_m and k_cat/K_m values, being read from their papers. How it works →Loading CatLog snapshot

Filters

Loading records

Export results

								Open record

Records per page:

CC BY 4.0How to cite

Page 1

CatLog in numbers

People behind CatLog

CatLog reads enzyme kinetics from the original papers, mostly those the enzyme databases cite. AI agents read and check each value; people review samples, and what they find improves the rules.

The loop runs clockwise from cited papers. Point at or tab to a person or an agent to see where they work.The list says where each person and agent works.

Using CatLog

CatLog reads enzyme kinetics (k_cat, K_m, k_cat/K_m) from the original papers, mostly those the enzyme databases cite. AI agents read and check each value; people review samples.

By the Chowdhury Lab, Iowa State University.

Records and values

Enzyme Organism Substrate k_cat s⁻¹ K_m mM k_cat/K_m mM⁻¹ s⁻¹ Review result 4-hydroxyacetophenone monooxygenase (HAPMO) Pseudomonas fluorescens 4‑aminoacetophenone 12.7 Acceptedfrom paper

One record and its three values, from Table 1 of one paper (PMID 12514023).

Record: One row of Browse: one enzyme, wild type or a variant, with one substrate under one set of assay conditions in one paper.
Value: One number: a k_cat, a K_m or a k_cat/K_m. A record holds up to three.
Paper: One publication; a paper can give many records.

Review result

Accepted: Read from the paper by CatLog; or, for a record not yet re-read, accepted by the earlier automated review.
Not confirmed: The earlier review left a question open.
Not checked: No review result. The record says Unverified (no final check) or Not reviewed (no PMID or DOI).
Disputed: A check found a problem.

How CatLog works

How CatLog reads a value

Read An AI model (Claude, Anthropic) reads each value with its unit, conditions and printed line, without seeing what the databases list.
Check A second, separate AI check compares every value with the paper. Automated checks confirm the line is in the paper.
Compare Each value is compared automatically with what the enzyme databases list for that paper.
Review AI audits and people re-check samples for systematic errors; what they find improves the next readings.

Units are converted automatically; no model types a converted number.

Records not yet re-read

CatLog’s earlier version built them from the enzyme databases. Its automated review checked part of them against their papers; each record’s review result says what it found.

Checks so far on Stats People behind CatLog

Read a record

Select a row to open its record.

4-hydroxyacetophenone monooxygenase (HAPMO)

EC 1.14.13.84 · Pseudomonas fluorescens · Wild type

Substrate: 4-aminoacetophenone

Acceptedhuman-checkedfrom paper

Values

Printed in Table 1
Parameter	Value	Unit	As printed
k_cat	12.7	s⁻¹	12.7 (0.7) s⁻¹
K_m		mM	0.82 (0.1) µM
k_cat/K_m		mM⁻¹ s⁻¹	15,000 × 10³ s⁻¹·M⁻¹

In the paper: “4-Aminoacetophenone 0.82 (0.1) 12.7 (0.7) 15,000”

Checks

Second check All values confirmed.

Human review Approved by a human reviewer.

ConditionsProtein and substratePaper

Header Enzyme, EC number, organism, form and substrate.
Review result The result and its tags (labels).
Values In catalog units, beside the cell as printed; a tick means a reviewer approved it.
In the paper The line the values come from.
Compared with databases What the databases list; here BRENDA’s K_m is 10× the paper’s (results).
Checks The second AI check, AI audits and human review (How CatLog reads a value).
Details Conditions, protein, substrate and paper.

Marks and cautions

Marks beside a value

Green tick: A human reviewer approved this value.
Exclamation mark: A caution; the record says what it is (Cautions).
calc.: k_cat ÷ K_m, worked out by CatLog.
—: No value (table cells only).

Cautions

What each beside a value means

Filter group	Cautions, as shown beside a value
k_cat/K_m mismatch	k_cat/K_m mismatch (the printed k_cat/K_m and k_cat ÷ K_m differ; all three are kept, and a factor near 60 points to minutes, near 1,000 to a concentration unit)
Unit or magnitude error	The unit or the value may be wrong. K_m > 10 M, K_m < 1 nM k_cat > 10⁸ s⁻¹ (the fastest known enzymes, such as catalase, reach a few × 10⁷ s⁻¹) possible 1000× unit error (the k_cat/K_m may be 1,000-fold off through its concentration unit) above diffusion limit (k_cat/K_m above 10¹⁰ M⁻¹ s⁻¹, 10⁷ mM⁻¹ s⁻¹) printed unit differs (from the concentration unit CatLog read)
No unit given	no unit (the source gives none)
Temperature questionable	T questionable (exactly 0 °C may mean none was given)

Evidence labels

On records not yet re-read

Located in the paper: Where the paper prints the values; the text may be context, not this value.
Note from BRENDA: The database’s own note; no paper text.
Several database records: Merged from several database listings; no paper text.
Paper cited: The paper is linked; where it prints the values was not saved.
Database record only: Only the database record.

Before you use the values

Only Accepted records passed review.
Read each value’s unit (Units).
A printed k_cat/K_m can differ from k_cat ÷ K_m (mismatch).
For training sets, split by paper or UniProt accession.

Reference

Find records

Search, filters and links

Search: An enzyme, EC number, organism, substrate, UniProt accession, PMID or DOI: laccase 1.10.3.2 P11509 12514023. A substrate word matches the start of a part of the name, so lactose does not find galactose. Gene names are not indexed.
Filters: Enzyme, EC number, organism, substrate; parameter; review result.
More filters: Source, wild type or variant, assay conditions, cautions and source database; it opens by itself when one is set.
Share: Every view and open record has its own address.

Try

Human-checked records
EC 1.1.1, wild type, 25–37 °C
Accepted k_cat/K_m without a caution, highest first

Units

k_cat s⁻¹ · K_m mM · k_cat/K_m mM⁻¹ s⁻¹

k_cats⁻¹Turnover number
K_mmMMichaelis constant
k_cat/K_mmM⁻¹ s⁻¹Catalytic efficiency (× 1,000 for M⁻¹ s⁻¹)

× 10ⁿ: From 100,000, below 0.001, and from 1,000 when trailing zeros are not figures (3.5 × 10⁴, not 35000).
Digits: Values read from the paper keep its significant figures; the downloads keep every stored digit.
T and pH: In the record and the downloads, not as table columns.
Other units: A few records not yet re-read keep another unit (min⁻¹, mg/mL) or no unit.

Downloads

Full data, the table and the comparison

Full data: Every record and field, with sequences, SMILES and paper links.
Table: The table’s columns with pH and temperature; joins Full data on record_key.
Compared with databases: One line per value and database, with the result and the database’s number (gzip CSV). Download the comparison
Format: gzip JSON Lines, one record per line: pandas.read_json(path, lines=True, compression="gzip"). The Dataset notes list every field.
Review files: Behind records not yet re-read: request by email.

Download full data

License and citation

CatLog’s compilation is free to reuse under CC BY 4.0, with credit. Quoted paper text remains the publisher’s; database records keep their own license.

Snapshot

—

Data ID

—

Cite

Sajeevan, K. A., et al. (2026). CatRange enables robust prediction of enzyme variant kinetic regimes. PNAS Nexus, pgag309. doi: 10.1093/pnasnexus/pgag309.

CatLog snapshot (date and data ID appear when the catalog loads)

More on citing

One value, the Data ID, CatLog-27k

For one value, add its record ID and keep the file you used: IDs of records read from the paper can change, and older snapshots leave this site.

The Data ID is the first 10 characters of the SHA-256 of the decompressed Full data file (gunzip -c catlog-enriched.jsonl.gz | shasum -a 256).

CatLog grew out of CatLog-27k, the CatRange paper’s training set.

Sources and their licenses

Cite any database whose records you use

Read from the paper: CatLog compilation CC BY 4.0; quoted paper text remains the publisher’s
Earlier paper extractions: As published (see paper)
BRENDA: CC BY 4.0
SABIO-RK: See the source’s terms
SKiD: CC BY-NC-ND 4.0
Open Enzyme Database (OED): CC BY 4.0
UniProt: CC BY 4.0

For reuse of CatLog review notes and corrections, contact the Chowdhury Lab.

Questions

Why does a value differ from BRENDA?

BRENDA may use another unit, round the value or take it from another table, condition or enzyme form.

Why is a sequence or SMILES missing?

CatLog adds one only when the enzyme matches one UniProt entry, or the substrate one structure; otherwise the field stays empty (uniprot_candidate_ids lists the candidates).

How do I report an error?

Choose Report a problem in the record. For anything else, use Report a data issue.