20/15 Visioneers and StableBody launch beta for enzyme expression-risk screening

20/15 Visioneers and StableBody Technologies have launched a beta program for industrial scientists to test a pre-synthesis screening platform that flags expression risks in enzyme designs. The companies will discuss the technology at Paperless Lab Academy USA 2026 in Cambridge, Mass., as they position the tool as decision support for costly recombinant expression failures.

Why it matters: - Recombinant expression is often a costly failure point in industrial enzyme development. - The new platform is meant to screen expression risk before gene synthesis, which could help teams avoid spending on constructs that are likely to fail in common production hosts. - The companies are positioning the tool as pre-synthesis decision support, not a replacement for experimental expression testing.

What happened: - 20/15 Visioneers and StableBody Technologies announced a partnership to commercialize computational tools for industrial enzyme design. - The companies also opened a beta program for industry scientists working on enzyme production. - Kerry Robinson, PhD, senior business development visioneer, and Kelly Cycon, PhD, life sciences director visioneer, will discuss the platform at Paperless Lab Academy USA 2026. - The conference runs Oct. 12-14 at The Charles Hotel in Cambridge, Mass. - Interested scientists can contact Kerry Robinson at kerry.robinson@20visioneers15.com or Kelly Cycon at kelly.cycon@20visioneers15.com.

The details: - The platform evaluates ribosome stalling, codon rarity transitions, co-translational folding bottlenecks, aggregation patches, and proteolysis risk in one analysis run. - It accepts DNA, protein, FASTA, and PDB inputs. - The system is parameterized for major industrial hosts and native sources. - It returns component-level diagnostics and prioritized variant designs. - Core analyses run locally and do not require paid third-party annotation services. - The solubility module was calibrated on experimental eSOL solubility using a held-out split of 294 examples, with Spearman rho around 0.55. - The module was then tested on a collaborator wet-lab set of 38 labeled constructs held out from model fitting. - On that set, the platform returned 68.4% accuracy, 56.5% sensitivity, 86.7% specificity, and MCC of 0.43. - The companies describe the result as useful triage signal on insoluble calls, not titer prediction. - The companies say the test set is small and project-enriched rather than an independent benchmark, and the confidence intervals are wide. - The co-translational modules have not yet been tested against external wet-lab outcomes. - The work appears in a preprint co-authored with Nicholas R. Silvaggi of the University of Wisconsin-Milwaukee, whose laboratory contributed the wet-lab constructs and outcomes. - The preprint is available at the preprint DOI and has not been peer reviewed. - The University of Wisconsin-Milwaukee does not endorse any commercial product. - Participants in the beta program submit candidate sequences under written confidentiality terms, receive full platform reports, and work with the joint scientific team on interpretation. - In return, participants are asked to share expression outcomes when agreements allow, so the model can be evaluated against host-matched industrial data. - Program terms, including sequence ownership and any use of shared outcomes for recalibration, are set out in the participation agreement. - Reports carry no warranty of expression outcome. - The beta program opens on [START DATE] and is limited to [COHORT SIZE] organizations in the first cohort.

Between the lines: - The disclosure is unusually direct about the platform’s limits, which signals that the companies are trying to build credibility with technical users rather than oversell early performance. - The focus on local execution and avoiding paid annotation services suggests an emphasis on practical workflow fit, not only model output. - The beta program structure indicates the companies want customer data to refine the system, especially for host-specific industrial expression cases. - The mention that preprint authors are employees or founders of the companies commercializing the platform underscores that the evidence base is still closely tied to the product team.

What's next: - The companies will use the beta program to test the platform on industry scientists' own constructs. - Participating teams are expected to return expression outcomes where permitted, giving the joint team additional data for evaluation and possible recalibration. - The platform will be further discussed at Paperless Lab Academy USA 2026 during the Oct. 12-14 meeting in Cambridge. - A key near-term question is whether the tool holds up across broader industrial datasets and different host systems.

The bottom line: - 20/15 Visioneers and StableBody are betting that an early expression-risk screen can cut wasted work in enzyme programs, even as they acknowledge the current evidence is still early and limited.