Generates and prioritizes protein sequences for design campaigns, including binder discovery.

DSM

DSM is Synthyra's protein generation research direction based on masked diffusion. It is designed to help researchers explore plausible protein sequence space and prioritize candidates for experimental follow-up.

What It Does

DSM supports:

• Protein sequence generation.
• Variant exploration around known templates.
• Binder design campaigns.
• Candidate triage before synthesis.
• Integration with Atlas scoring, annotation, and structure checks.

Why It Matters

Protein design is a search problem. The number of possible sequences is enormous, while the number that can be tested in the lab is small.

DSM helps make that search more directed. It generates plausible candidates and works with other Synthyra tools to rank which designs deserve attention.

Research Foundation

The DSM paper showed that masked diffusion can support both protein representation and generation. In experimental validation, DSM-generated EGFR and PD-L1 candidates produced real expression and binding signals, including a sub-nanomolar EGFR binder.

Synthyra's product direction extends this research with validation workflows, platform integration, and design triage layers.

Intended Use

Use DSM for early design exploration, especially when a team needs candidate sequences to rank, inspect, and validate. It is best used with Atlas PPI, structure prediction, annotation tools, and wet-lab follow-up.

Limitations

DSM does not guarantee expression, folding, binding, safety, or function. Generated proteins can fail for many reasons outside the model's view. All consequential designs require experimental validation and responsible review.