Skip to main content

AI-guided peptide design

Unlocking the Full Therapeutic Potential of Peptides

We are building a platform to advance promising peptide therapeutics, design entirely new molecules, and help partners address longstanding barriers to clinical development.

GLP-1 binds to GLP-1R. This public reference complex uses experimentally resolved atomic coordinates from PDB 6X18. The animated approach is illustrative. Unresolved residues and auxiliary proteins are omitted. The peptide is highlighted through the receptor surface to show the binding site.
From molecular recognition to therapeutic possibility.Our vision

Our vision

Peptides are among nature’s most extraordinary molecules.

Peptaris is building an AI-powered platform to design peptides with the properties they need to become viable medicines.

The design–make–test loopA peptide lead enters a continuous loop of design, score, synthesize and test steps around the ApexChem Model Suite. Measured test results feed back into the model for the next round.PEPTIDE LEADDESIGNSCORESYNTHESIZETESTAPEXCHEMMODEL SUITE
The design–make–test loop
  1. Balance Multiple Properties at Once

    We generate optimized candidates with targeted functional and developability properties.

  2. Laboratory Data Generation

    Our high-quality synthesis and experimental data libraries are architected for reliable model training.

  3. A Continuous Learning System

    The sophistication and accuracy of our molecular designs compound with every cycle.

The opportunity

The Next Frontier in Peptide Therapeutics

Peptides are among biology’s most versatile molecules, playing essential roles in many of the body’s most important functions. They have already given rise to important medicines, yet much of their therapeutic potential remains unrealized.

Many promising peptides are unstable, short-lived, or difficult to deliver effectively. Advances in artificial intelligence and medicinal chemistry are creating new opportunities to overcome these challenges and expand what peptides can achieve.

  • Advancing Promising Peptides

    Addressing the properties that have limited the development of biologically promising peptides, helping move them closer to viable therapeutic compounds.

  • Reimagining Existing Successes

    Engineering known peptides to achieve a more balanced set of pharmacological properties and create novel molecular designs.

  • Designing New Molecules

    Exploring entirely new peptide designs with combinatorial chemistry to expand possibilities.

Lineage

Each generation of peptide medicine solved one more problem.

The next one has to solve many at once.

  1. GEN.01

    Natural peptides

    Hormones and host-defence peptides evolved for function, not for stability or delivery.

    insulin · oxytocin · AMPs

  2. GEN.02

    Engineered analogues

    Medicinal chemistry tunes one property at a time: half-life, potency, or selectivity.

    octreotide · desmopressin

  3. GEN.03

    Modern peptide therapeutics

    Lipidation and backbone changes produce blockbuster medicines, proving what peptides can do.

    GLP-1 receptor agonists

  4. GEN.04

    AI-designed peptides

    Models trained on purpose-built lab data balance many functional and developability properties at once.

    Peptaris

Our approach

A New Era of Peptide Design

Peptaris is building an AI-powered platform to design peptides with the properties they need to become viable medicines.

A scientist working on peptide chemistry in the laboratory
Experimental peptide chemistry
Model development displayed on a computer screen
Model development

Our foundation

Built on Pioneering Science

Peptaris was founded by scientists from the University of Pennsylvania’s Machine Biology Group and entrepreneurs with a shared ambition to advance the field of peptide design.

Building on research in AI-guided peptide discovery, our team brings together expertise in predicting how peptide sequences influence biological activity and engineering molecules with improved therapeutic properties.

Our ambition is to translate advances in molecular design and medicinal chemistry into the next generation of peptide medicines.

In the news

News & Insights

Discover the latest news from Peptaris and the pioneering research of our scientific founders.

Careers

Join Our Team

We are building an interdisciplinary team of scientists and engineers to advance the next generation of peptide design.

Join us at the intersection of artificial intelligence, chemistry, and biology, where ambitious ideas become scientific breakthroughs.

FAQ

Frequently asked questions

What does Peptaris do?

Peptaris is building an AI-powered platform to design peptides with the properties they need to become viable medicines: advancing promising peptides, reimagining existing successes, and designing entirely new molecules.

Why peptides?

Peptides are among biology's most versatile molecules and have already given rise to important medicines. Many promising peptides, however, are unstable, short-lived, or difficult to deliver. Advances in AI and medicinal chemistry are creating new opportunities to overcome these challenges.

How is the platform different?

We optimize multiple functional and developability properties at once, generate our own high-quality laboratory data architected for model training, and run a continuous learning loop in which every design cycle improves the next.

Are the publications on this site studies conducted by Peptaris?

No. The publications we list describe academic research and collaborative contributions from our scientific founders and their collaborators. They inform our scientific foundations but are not evidence of the current performance of our commercial platform.

How can we partner with Peptaris?

We collaborate on existing and emerging therapeutic programs. Write to info@peptaris.com and tell us about your program.

Partnerships

Partner with us in advancing peptide science, and bringing its potential to the world.

Your Privacy Choices

Strictly necessary
Remembers this choice in your browser. Always on; it identifies no one.
Analytics (Vercel Web Analytics)
Cookieless, aggregated page-view statistics. Currently: Off