bioRxiv Science⌕ Search
Home/Archive/bioRxiv

bioRxiv · 10.64898/2026.06.22.733870

A high-level programming language for generative biology with Proto

Merchant, A.·Guo, D.·Viggiano, B.·Brennan-Almaraz, L.·Hurr, E.·Mai, T.·Yin, P.·King, S.·Ashley, E.·Hie, B.
View original source ↗Export metadata ↓

Abstract

Programmable composition of complex systems is a longstanding goal of biological research. Generative modeling has improved the reliability of computational design, but existing methods are highly specialized and are difficult to extend or compose. Here, we introduce Proto, a high-level programming language for generative biology. By composing a small set of abstract primitives into structured programs, Proto encodes generative design campaigns across diverse modalities and scales--spanning DNA, RNA, proteins, ligands, and their interactions. Proto readily incorporates predictive models into generative workflows, which we leveraged to design alternatively spliced introns with experimental validation in human cell lines. Proto is natively multi-objective, enabling the design of promoter-repressor pairs with leading experimental success rates for synthetic protein-DNA design. Alongside AI agents, Proto enables the specification of complex pathways and regulatory logic through natural language instructions. We openly release Proto, including software infrastructure and user interfaces, to enable widespread access to generative biological programming.

Explore related subjects

Synthetic BiologyPreprint

Keep this discovery

Explore connections, maps & timelines

BibTeXRIS

Merchant, A., Guo, D., Viggiano, B., Brennan-Almaraz, L., Hurr, E., Mai, T., Yin, P., King, S., Ashley, E., Hie, B.. 2026-06-23. A high-level programming language for generative biology with Proto. https://doi.org/10.64898/2026.06.22.733870

Cite the original work for its findings. Save a collection to share your selection of sources.

KEEP EXPLORING

Related preprints

bioRxiv2026

AI-enabled rhodopsin design for blue-light enhanced bacterial growth

We developed an AI-guided design pipeline that generated and validated non-natural microbial rhodopsins with spectral properties not yet known in nature. The pipeline comprised a three-stage in silico design, a genetic algorithm (GA) for sequence generation, a stacked LASSO and XGBoost machine-learning (ML) regressor for spectral prediction and fitness ranking, and a Markov-based sequence plausibility filter to enforce proton pumping like characteristics. Four candidate rhodopsins (APR1, APR2, APR6, and APR7) targeting blue light absorption were designed and AlphaFold3 structural modelling predicted retinal binding pocket architecture consistent with outward proton-pumping function. Experimental characterisation confirmed that all four variants absorbed light at [~]410 nm and significantly promoted the growth of Cupriavidus necator under blue light illumination. This study demonstrates that AI-enabled design can engineer proteins with no natural precedent, generating light-harvesting rhodopsins with novel spectral properties while preserving biological function, marking a significant advance in programmable synthetic biology.

Synthetic Biology↗
bioRxiv2026

Systematic dissection of Cas12a-mediated precision genome editing defines design principles for genome-scale variant engineering

Cas9 precision editing is increasingly predictable because guide, donor and target-context effects have been systematically characterized. Extending this framework to other nucleases is essential for installing variants outside convenient Cas9 target space. Cas12a provides a T-rich protospacer-adjacent motif (PAM) alternative, but determinants of efficient donor-templated Cas12a editing remain poorly defined. Here, we systematically dissected Cas12a precision editing in Saccharomyces cerevisiae across nuclease, direct repeat, expression, crRNA, donor, genomic context and time-course variables. Reporter and amplicon-sequencing assays showed that cleavage activity alone did not predict precise editing. Highly active configurations often reduced viability or lost edited alleles over time, whereas attenuated configurations better preserved programmed edits. Enhanced AsCas12a edited rapidly and tolerated shorter crRNAs, resulting in a narrower editing window, while an attenuated FnCas12a configuration edited more slowly but maintained higher viability and better distal-edit recovery. Alternative repair outcomes were rare, target-dependent, and further suppressed by LexA-FHA donor recruitment. To define design parameters at scale, we established a pooled Cas12a platform with 530 barcoded edit cassettes and recovered programmed edits for 70.2% of designs. Successful editing was reduced with TTTG PAMs, a C upstream of the PAM and at distal edit positions. Excluding these features increased the edited fraction to 85.4% and adding high predicted cleavage scores further elevated it to 91.4%. Applied retrospectively, these criteria also identified poorly edited loci in the targeted panels. Together, these data define design principles for Cas12a-mediated precision editing and establish a scalable platform for genome-scale pooled variant engineering and phenotyping in yeast.

Synthetic Biology↗
bioRxiv2026

A Minimal Packaging Signal Enables Production of High-Purity Phage-Like Particles for CRISPR-Cas Antimicrobials in Staphylococcus aureus

Precision phage therapeutics provide a promising strategy to combat multidrug-resistant pathogens, including Staphylococcus aureus. Efficient, specific packaging of genetic cargoes remains challenging. Using modular design principles, we report a minimal phagemid packaging signal consisting of the phage terminase small subunit under its native promoter that significantly outperforms conventional packaging signals. The utility of this synthetic terS operon was demonstrated through production of highly concentrated and genetically pure CRISPR-Cas antimicrobials. To circumvent CRISPR-mediated self-targeting during antimicrobial generation, a terS-deficient strain was engineered to express the anti-CRISPR protein AcrIIA4, enabling titers above 1010 transducing units per milliliter (TRU/mL) with over 94% purity. With a high-copy origin of replication module, CRISPR-Cas phage-like particle titers could approach 1012 TRU/mL. We discovered that pure CRISPR-Cas antimicrobials are potent and can be amplified in hosts possessing prophages. Taken altogether, this study defines the minimal and optimal genetic requirements for efficient, specific creation of phage-based technologies. Technological ReadinessThe described system for engineering phage-like particles has reached a technological readiness level (TRL) of 4-5 based on the provided laboratory validation and strong literature support from other engineered phage therapies applied to in vivo models. Previous systems have demonstrated high-purity phage-like particle preparation, but always at the cost of severely reduced productivity. Therefore, our demarcation of a specific, efficient, and minimal system for packaging nucleic acid cargoes into phage vectors is a critical step toward real-world use. Despite this, low levels of contaminating host/phage DNA remain a key barrier to phage-based therapies. Protein and strain engineering efforts can help mitigate terminase nonspecificity, but care must be taken to not compromise productivity. More generally, widespread adoption will require deeper understanding of host-pathogen-phage interactions, development of scalable GMP manufacturing processes, and harmonized regulatory guidance that recognizes the dynamic nature of phage-derived technologies.

Synthetic Biology↗