KaML: Acid-Base Constant Prediction
From Protein Sequence (ESM) or Structure (CBT)

Usage Notice & License

KaML is provided exclusively for non-commercial academic and research purposes. By using this tool, you agree to its licensing under the PolyForm Noncommercial License 1.0.0.

If you use KaML in your work, please cite:

M Shen, GW Dayhoff II, D Kortzak & J Shen. Protein Electrostatic Properties are Fine-Tuned Through Evolution  In revision. bioRxiv.

M Shen, D Kortzak, S Ambrozak, S Bhatnagar, I Buchanan, R Liu, & J Shen. KaMLs for Predicting Protein pK_a Values and Ionization States: Are Trees All You Need?   J Chem Theory Comput 2025.

Built with ESM

This tool is powered by Evolutionary Scale Modeling (ESM), utilizing the following models:

• ESM-3 (esm3-large-2024-03)
• ESM-C (esmc-6b-2024-12)
• ESM-2 (esm2_t33_650M_UR50D)

If you don’t have an ESM forge API token, please visit forge.evolutionaryscale.ai/sign-up. After you sign up, visit the 'Console' page to create a free ESM forge API token*.

*ESM forge API tokens are used solely by KaML for executing predictions and are never stored or tracked.

The ESM safety filter is designed to prevent generation of outputs that may be unsafe. By choosing to skip the filter, you acknowledge that you have the appropriate permissions and fully understand the risks associated with bypassing this safety mechanism.

---

Here is the language provided by ESM regarding so-called "sequences of concern" taken from https://forge.evolutionaryscale.ai/faq on 4-8-2025:

"We implement a guardrail which will detect sequences and keywords of concern at the input or output of ESM3 API calls, to prevent work on harmful sequences. However, we recognize that there are many legitimate reasons to use AI models to understand and model these sequences (e.g. vaccine development or detecting new variants of concern). If you want to use ESM3 for such use cases, please contact us on info@evolutionaryscale.ai with subject line “Request for Elevated Access.” Your request should contain your name, institutional affiliation, and a description ofthe use-case that you are requesting elevated access for."