Property prediction · Cheminformatics

CatPred API

Predict kcat, Km, or Ki for an enzyme and substrate, with an uncertainty estimate.

Runs CatPred, which combines a pretrained protein language model with a molecular representation of the substrate and predicts a distribution rather than a point value, so each prediction carries a variance that tracks how far the query sits from the training data.

Web form Documentation ↗ Home page ↗ Source code ↗

Submit POST /api/catpred/

Queues a run and returns 202 with the job.

Schema GET /api/catpred/

Returns this tool's metadata and the fields below, as JSON.

Runtime

Installed by setup_system.sh into its own uv environment; the checkpoint archive downloads automatically, and setup_system.sh symlinks one ensemble per parameter (seed0, picked arbitrarily among the ten this research-reproduction bundle ships) into a consistent kcat/km/ki layout under CATPRED_CHECKPOINT_DIR. The ki checkpoint needs pretrained EGNN features this adapter does not supply, so that parameter is not fully wired up · Python app (uv) · Expensive (min or hours)

Request body

Fields

The same names the web form posts. See the field type table for what each kind means over HTTP.

Name	Type	Required	Default	Description
`job_name`	string text	no	`catpred-demo`	Job name
`parameter`	string select	no	`kcat`	Parameter One of: `kcat`, `km`, `ki`.
`protein_sequence`	string textarea	yes	`MSEPAQKKQKVSNGGGSSNKRAADSLADQLYGKSAAAAHTPPAKKAKTETIAAPTFTTSGLDKLDLSNIDSSAWKQLAEEDLASLYGDLD …`	Enzyme sequence
`substrate_smiles`	string textarea	yes	`OCC1OC(O)C(O)C(O)C1O`	Substrate SMILES For a Ki prediction this is the inhibitor rather than the substrate.
`ec_number`	string text	no	`—`	EC number (optional) For example 2.7.1.1. Used as a feature where the model has one. up to 20 characters.
`organism`	string text	no	`—`	Organism (optional) For example Escherichia coli. up to 200 characters.
`include_uncertainty`	boolean checkbox	no	`true`	Return the predicted uncertainty Lower predicted variance goes with higher accuracy, so this is worth keeping on.

Example

A request that runs

These are the defaults, exactly as the web form would post them.

curl -X POST https://www.athanortools.com/api/catpred/ \
  -H 'Content-Type: application/json' \
  -d '{
  "job_name": "catpred-demo",
  "parameter": "kcat",
  "protein_sequence": "MSEPAQKKQKVSNGGGSSNKRAADSLADQLYGKSAAAAHTPPAKKAKTETIAAPTFTTSGLDKLDLSNIDSSAWKQLAEEDLASLYGDLDSHRLDQPFPSAAAPVAKKRVSFTDSAAAA",
  "substrate_smiles": "OCC1OC(O)C(O)C(O)C1O",
  "ec_number": "",
  "organism": "",
  "include_uncertainty": true
}'

The reply is 202 with a queued job; poll its status_url until status is succeeded or failed. See the quick start for the whole exchange.

Responses

What comes back

status	Meaning
`queued`	Accepted, waiting for the jobs ahead of it. `position` counts how many those are.
`running`	The tool is executing now.
`succeeded`	Finished; `result` holds the tool's output and `license` the terms it came under.
`failed`	Finished; `error` holds a code and a message.

Errors

code	Meaning
`invalid_input`	The client supplied invalid or incomplete input.
`tool_unavailable`	The requested third-party dependency is not available on this host.
`execution_failed`	A configured third-party process exited unsuccessfully.
`internal_error`	An adapter failed in a way it does not describe. The detail is in the server log, not the response.
`not_found`	No job has that id. Finished jobs are dropped eventually.