Discovery platform

Design, make, test. Learn faster.

We are building a closed-loop discovery platform where AI-designed libraries meet phage display, and every experiment makes the next library smarter.

Platform in development: enquire about early collaborations
The discovery cycle

From virtual hit to optimised lead.

Each cycle supplies the evidence needed to make the next library more focused and more likely to contain high-quality binders.

  1. AI library design

    Computational analysis of the binding site guides AI and machine-learning models to predict high-affinity sequences before wet-lab work begins.

  2. Phage display panning

    The enriched library is displayed on phage and taken through three to five rounds of selection. Binders are eluted and amplified.

  3. Hit identification

    Enriched clones are sequenced by NGS. Binding is confirmed by ELISA and SPR, and the data is fed back to the model.

  4. Lead optimisation

    SAR, alanine scanning, cyclisation and stapling improve potency, selectivity and stability.

How the loop works

Real binding data makes the next cycle smarter.

Diagram of the SMS closed-loop peptide discovery platform: AI library design, phage display, hit identification and lead optimisation
  • 01
    Confirmed hits teach the model

    Binding, sequence and structural data from real hits go back into the AI.

  • 02
    Non-binders matter too

    Sequences that fail help the model filter out poor candidates.

  • 03
    Libraries are redesigned

    The model builds the next library using both binder and non-binder data.

  • 04
    Each library is more focused

    Libraries become progressively enriched for your target.

  • 05
    Hit rates climb

    Each round of phage display is designed to deliver a higher hit rate than the last.

Why it works

A compounding advantage for discovery teams.

Compounding learning

Hit quality improves with every experimental cycle, not just the first.

Grounded in reality

Libraries are built around what actually binds, not only on predictions.

Fewer wasted cycles

Fewer panning rounds can be needed to surface strong candidates.

Best hits rise faster

Promising candidates are prioritised earlier, shortening timelines.

Discovery capabilities

Move from target to validated hit.

Computational design

Use target information and computational analysis to prioritise focused peptide libraries.

Phage display

Apply phage display to identify binders against specific targets.

Hit confirmation

Synthesise and characterise selected sequences to support hit-to-lead decisions.

Program progression

Hand confirmed hits straight to our hit-to-lead and lead optimisation team.

Have a target in mind?

Talk to us about early collaboration on our discovery platform.

Start a conversation →
business@smspeptide.com +91 40 2525 9999 LinkedIn