# Biomatter raises $18 million for AI enzyme design

> The company plans a new facility to expand design and testing.

By BIGCHANGE.AI Editorial

Published: 2026-11-12T10:47:38.473Z
Updated: 2026-11-12T10:47:38.473Z
Canonical: https://bigchange.ai/blog/biomatter-series-a-ai-enzyme-design

![Laboratory testing.](https://bigchange.ai/api/media/file/enzyme-lab-concept.png)
AI-generated conceptual illustration for BIG CHANGE

Biomatter has raised $18 million for AI enzyme design, including a Vilnius testing facility planned for 2027.

## The big change

- Biomatter says it designs enzymes for a customer's desired reaction, then tests them. Enzymes are proteins that help chemical reactions happen; a proposed sequence still needs experimental verification.
- The company wants to shorten development from weeks to days. That is a goal, not an independently measured improvement.
- Earlier research demonstrated some working generated enzymes. Its laboratory results explain why testing matters, without establishing a success rate for today's commercial platform.

## From a sequence to a working enzyme

Biomatter's [announcement](https://investor.wedbush.com/wedbush/article/bizwire-2026-11-12-biomatter-raises-18m-series-a-to-scale-its-ai-powered-enzyme-design-for-global-industry-leaders) says Astanor led the financing and the new facility will bring more testing in-house.

A protein's sequence is the order of its building blocks. Generating that sequence is only one step: researchers must make the protein and test whether it performs the intended reaction. A plausible design alone cannot establish performance in a customer's process.

A [2021 Nature Machine Intelligence paper](https://www.nature.com/articles/s42256-021-00310-5) described ProteinGAN, which learned from natural protein sequences to generate new variants. Researchers used malate dehydrogenase, or MDH, as the template enzyme. Of 55 generated sequences tested, 13 were soluble and showed MDH catalytic activity under the laboratory conditions used.

Solubility and activity are separate questions: can the protein remain dissolved, and can it carry out the reaction? The reported 13 met both criteria. This was evidence of functioning generated proteins in a particular experiment, not a general 24% success rate for enzyme design. It does not establish the performance of the current commercial system or explain every unsuccessful candidate.

## What a commercial example establishes

Biomatter says it helped improve oxidation resistance in Thermo Fisher Scientific's Invitrogen SuperScript V reverse transcriptase. The manufacturer's pages do not identify Biomatter's contribution, so that attribution remains the developer's claim.

A reverse transcriptase makes DNA from an RNA template. This gives the product a concrete research role: it helps laboratories convert material from one molecular form into another for subsequent work.

[Thermo Fisher's overview](https://www.thermofisher.com/us/en/home/life-science/pcr/reverse-transcription/superscript-reverse-transcriptase.html) lists room-temperature storage of up to six months for the stand-alone enzyme, compared with one month for master mixes. Those are different formats, so the longer period should not be applied to every product carrying the name. Its [catalog listing](https://www.thermofisher.com/order/catalog/product/19090010) specifies research use only, not diagnostic procedures.

Storage specifications matter for laboratory handling, but they answer a different question from whether a reagent is suitable for diagnosis. Buyers need to check both the format and the permitted use rather than treating improved stability as broader authorization.

For industrial customers, useful acceptance tests would address the intended reaction, operating conditions and production scale. A successful small experiment cannot settle all three. The practical question is what evidence supports the particular application, rather than whether AI was used to propose the sequence.

## Sources

- [Biomatter funding announcement](https://investor.wedbush.com/wedbush/article/bizwire-2026-11-12-biomatter-raises-18m-series-a-to-scale-its-ai-powered-enzyme-design-for-global-industry-leaders) — Funding and developer claims.
- [ProteinGAN research paper](https://www.nature.com/articles/s42256-021-00310-5) — Earlier generated-enzyme experiment.
- [Thermo Fisher SuperScript overview](https://www.thermofisher.com/us/en/home/life-science/pcr/reverse-transcription/superscript-reverse-transcriptase.html) — Product formats and storage.
- [Thermo Fisher product listing](https://www.thermofisher.com/order/catalog/product/19090010) — Research-only use and storage.