Every engagement starts with a free read on published data, so you see the report format before you pay for one. A paid single-compound mechanism read starts at $12,000 and takes one to two weeks.
The mechanism answer and the experiment that confirms it are what you buy; the risk label comes along.
One compound, from ion-channel data that is already published or that you are free to share. You get the same style of report you would later pay for, so you can judge the work on its substance.
Ranked ion-current mechanisms that reproduce the signal
Fractional block per current from 1 to 4x free Cmax, and the APD90 change
The TdP-risk class those currents map to
The confirming experiment for that compound
The limits of the read, stated in full
It is never invoiced. Confirming-experiment design on a proprietary compound is paid work.
Your compound and your data, run on the Kernik-Clancy 2019 human iPSC-cardiomyocyte model. Send a standard ion-channel panel, or a clear repolarization signal. The read is only as wide as the channels you send, and every other input is fixed and published, so the claim is checkable in your own lab.
Everything in the first read, on your own data
Competing mechanisms ranked, with the ones the data leave tied named as such
An identifiability analysis over the model conductances at your observed APD90
The single assay that discriminates the surviving mechanisms
A written report you can hand to a toxicologist
Price moves with the number of currents in the panel and the state of the data. It is quoted before any work starts.
On a proprietary compound the panel and the Cmax are yours to supply: a hERG patch-clamp IC50 and a free clinical Cmax, which is an assay and a Phase 1 result. The free hERG margin on the benchmark page is only free because someone else published both inputs.
After you submit your panel, we confirm the exact price and send a payment link or invoice. Work begins once payment is received.
For a chemical series, or for a group sending panels on a regular cadence. Priced per compound below the single-read rate, or as a monthly retainer with an agreed number of reads. Ask and we will scope it.
If you decide to run the confirming experiment, we design the iPSC-cardiomyocyte pilot around it: the arms, the concentrations, the readouts, and what each result would rule out. You commission it at a lab of your choosing, you keep the data, and we help you read it against the model.
This is optional and secondary. The read stands on its own. No wet pilot has run for a customer yet.
1. You send the data. A standard ion-channel panel, or a clear repolarization signal, plus the free Cmax or concentration range that matters for the decision. A short note on context helps, though it stays optional.
2. We confirm scope and price. You receive a clear quote before any work begins (typically starting at $12,000 for a single-compound read).
3. Payment. We send a Stripe payment link or a formal invoice. Work starts once payment or a purchase order is received.
4. Delivery. You receive the full written mechanism report within 1–2 weeks.
The free first read (published or non-confidential data only) uses the same report format, so you can evaluate the quality of the work before paying for a proprietary compound.
what you should know before you pay
The prices above are the ones we are testing. Perturb is early and this is a new offer, so the figures may move. Whatever is quoted to you before work starts is the price you pay.
The model is single-cell. It bounds the causes of a repolarization signal; the confirming experiment proves one. Conduction and QRS sit out of scope, and Kernik-2019 carries no distinct late-sodium current, so INaL block stays out of its view.
We publish the benchmark where we lose. On the public 28-drug CiPA panel, a free hERG-only margin still beats our three-class risk call. The benchmark page carries that number in full, along with the compounds the model misses. Read it before you buy.
What you are paying for is the mechanism and the next experiment, on the same cell type your confirmatory assay uses, so the prediction is checkable in your own lab. If a risk class alone is what you need, CiPA's own model is free and stronger at it, and the benchmark page shows the comparison.
One compound, from published or otherwise non-confidential channel data. Tell us the compound and what the panel shows, and we will tell you whether the read can say anything useful before we run it.