The company building drugs from the mirror image of nature - peptides made entirely of D-amino acids, designed by AI.
Here is a fact about your body that is both true and slightly unsettling: nearly every protein in it is left-handed. Amino acids come in two mirror-image forms, L and D, and biology - for reasons that trace back to some accident of early chemistry - picked L and never looked back. Aizen Therapeutics looked back. The San Diego startup, spun out of a Caltech lab, builds what it calls Mirror Peptides: therapeutic molecules assembled entirely from D-amino acids, the reflection of the ones nature uses.
This is not a gimmick, and the reason is the interesting part. Peptide drugs have two chronic, well-documented problems. The first is that your body's proteases - the enzymes that chew up proteins - are very good at their job, so a peptide drug often does not last long. The second is that the immune system can decide a peptide is a threat and attack it. Both problems share a root cause: the machinery that degrades and recognizes peptides is itself built to handle L-amino acids. Feed it a D-amino acid mirror image and it fumbles. The molecule it was built to grab no longer fits its hand.
So a full D-peptide can, in principle, bind its target with the potency of a biologic while shrugging off the proteases and slipping past the immune system. Chemists have known this for a long time. The catch has always been design: you cannot simply mirror a known drug and expect it to work, because the target it needs to bind is still left-handed. You have to design the mirror molecule from scratch, against a real, chiral target. That is a computational problem of enormous size, and it is precisely the problem Aizen built an AI to solve.
That AI is called DaX - the D-amino acid eXplorer - and it is the whole company in a sense. It emerged from the lab of Caltech professor David Van Valen, and it does something more specific than "generative AI for drugs." It trains foundation models to understand the physics of how receptors and ligands actually bind, then uses guided diffusion to design a D-peptide that fits a chosen target. Aizen says it has validated this across multiple clinically relevant targets. The pitch, printed plainly on its own website, is "medicines beyond nature."
The more D-amino acids a peptide contains, the more stable it is - proteases struggle to break down molecules they don't recognize.- The stability logic behind Mirror Peptides
D-amino acid eXplorer. A first-in-class computational protein-design engine that pairs generative AI with structural biology, learns receptor-ligand physics, and uses guided diffusion to design de novo D-peptides against real targets.
Synthetic peptides made entirely of D-amino acids - the mirror image of natural proteins. Engineered for strong affinity and specificity, enhanced stability, and reduced immunogenicity.
An internal set of drug candidates, including a focus on oral peptide therapeutics for chronic immune disorders such as cytokine inhibition. The ambition: a biologic you can swallow.
A repeat operator based in San Diego, Kshatriya previously founded the genomics diagnostics company Biota. He met Van Valen and set out to build a startup that pointed AI at drug discovery rather than diagnostics.
The scientific engine. DaX emerged from Van Valen's lab at the California Institute of Technology, where his work sits at the intersection of machine learning and biology. Aizen is the company that spun out of it.
Scientific advisor: Michael Kay (University of Utah School of Medicine), a D-peptide therapeutics expert.
| Round | Amount | Date | Lead & Investors |
|---|---|---|---|
| Seed | $13,000,000 | Nov 2024 | Madrona Venture Group (lead), Wilson Hill, Cercano Management |
When Aizen came out of stealth in November 2024, it led with the platform rather than a single disease. That is a deliberate wager: build the design engine first, and the drug candidates - internal and partnered - follow across many targets.
Aizen belongs to the fast-growing cohort of AI-driven protein and peptide design companies - names like Generate:Biomedicines, Cradle, and Chai Discovery - plus the broader world of established peptide and macrocycle drug developers. What sets Aizen apart is the chirality frontier: most players design within nature's L-amino-acid toolkit. Aizen is designing outside it, in the mirror.
9276 Scranton Rd, Suite 500, San Diego, CA 92121 · Aizen Therapeutics, Inc.
Aizen Therapeutics is a San Diego, AI-driven biotech spun out of Caltech that designs Mirror Peptides - synthetic drugs built entirely from D-amino acids, the mirror image of the L-amino acids that make up natural proteins. Its DaX platform pairs generative AI with structural biology to design these molecules from scratch, aiming for drugs with the potency of biologics but greater stability, lower immunogenicity, and the potential to be taken orally. The company emerged from stealth in November 2024 with a $13M seed round.
Last updated: