CAMOLE-1: ML-driven discovery of selective MMP12 inhibitors

bioRxiv USA

Overview

Calico Life Sciences developed CAMOLE-1, a machine learning platform that utilizes an all-atom diffusion model to design potent and selective peptide inhibitors. The system successfully optimized an MMP12 inhibitor over approximately one year through the synthesis of 270 peptides. This approach demonstrates the ability of ML to drive peptide design from initial hit discovery to lead development with improved pharmacokinetic properties.

In Depth

Key Findings

The CAMOLE-1 platform enabled the discovery and optimization of a potent and selective inhibitor for matrix metalloproteinase 12 (MMP12). By integrating an all-atom diffusion model with physics-based computational filtering and biochemical testing, the researchers developed a lead peptide inhibitor within approximately one year, synthesizing a total of 270 peptides during the process.

Technical Details

CAMOLE-1 is a de novo peptide inhibitor design platform. Its core technical components include:

• All-atom diffusion model: This enables the rapid discovery and optimization of non-canonical peptides.
• Iterative Workflow: The process began with canonical amino acid-based designs to identify an initial MMP12 peptide inhibitor, which was then further optimized for potency.
• Structural Validation: Predicted structures for both canonical and non-canonical peptide inhibitors showed high fidelity when compared to experimentally determined peptide-bound crystal structures.
• Integrated Filtering: The platform combines machine learning with physics-based computational filtering and biochemical testing to refine candidates.

Background & Context

Peptide therapeutics are an emerging modality offering increased potency and reduced toxicity compared to other options. The researchers targeted matrix metalloproteinases (MMPs), a class of targets that has essentially evaded therapeutic development until now. This work represents the first described effort to use machine learning techniques to iteratively drive peptide inhibitor design beyond the initial hit discovery phase into the development of selective, potent lead inhibitors.

Significance & Outlook

The successful application of CAMOLE-1 illustrates the potential of machine learning to accelerate drug discovery pipelines for complex targets. By improving pharmacokinetic properties and achieving high selectivity, the study highlights the power of ML-driven design in the emerging field of peptide therapeutics.

Get our weekly technology intelligence — free

Receive an infographic that lets you judge at a glance whether each field’s analysis report is worth reading.

Subscribe Free — Weekly Tech Intelligence

By subscribing, you’ll receive Troy-Technical’s weekly technology intelligence newsletter.

• Your email and selected fields are used only to deliver the newsletter.
• We never share your information with third parties.
• You can unsubscribe anytime via the link in each email.

See our Privacy Policy for details.

Agree & Continue ▶

Takes about a minute · Unsubscribe anytime