Overview
The next novel plant trait is waiting to be discovered and optimized
Through access to fast and affordable sequencing, companies and research organizations have amassed tons of biological data, connecting genes to specific structures and functions. This genomic data, combined with generative artificial intelligence, is poised to transform agriculture.
- Discover starting points for novel plant traits in our proprietary genomic database comprising 2B+ protein sequences
- Optimize leads with AI-driven protein and enzyme engineering geared towards improving efficacy and stability
- Screen candidates for disease or insect control, or herbicide tolerance using ultra-high throughput methods
- Optimize gene constructs to deliver lead candidate traits with greater precision and predictability
Solutions for
- Input trait discovery
- Protein and enzyme engineering
- Multi-gene construct design
- Codon optimization for reliable expression
- Promoter optimization for targeted expression
01 TRAIT DISCOVERY
Find novel, unencumbered proteins in our massive metagenomic library
Ginkgo has amassed metagenomic libraries spanning diverse sources. The ~2.7 billion genes in our library, sourced primarily from uncultured microbes, have <5% overlap with publicly available sequences.
Why is this important? For a few reasons – the likelihood of discovering a trait with novel mechanism of action and freedom to operate increases with the size of the library being mined. Furthermore, larger libraries offer more data to tune machine learning models designed to enrich and nominate novel and efficacious traits.
Case Study: Finding novel insect control proteins (ICPs) for new mechanisms of action (MoA)
The Challenge. A Tier 1 agricultural company with insect control protein traits on the market observed increasing pest resistance to its traits. While the company had a protein library to scan for trait candidates, the best performing leads in the library fell short of having freedom to operate or new mechanisms of action that would bypass resistance.
The Solution. Using a machine learning model that builds off of known insect control proteins and patented enrichment and amplification techniques, Ginkgo combed through its metagenomic database to nominate 220 candidate traits. Further refinement criteria narrowed this pool down to 92 putative proteins that Ginkgo produced and purified for diet bioassays. While our partner initially deemed one hit “a homerun”, Ginkgo found a total of three proteins that worked in the killing assay, showing potential new insect control mechanisms of action.
02 LEAD PROTEIN OPTIMIZATION
Improve the performance of your lead candidates with machine-learning guided protein optimization
Through >200 unique R&D programs, Ginkgo has run more than 400 enzyme improvement campaigns across a range of enzyme classes. These campaigns have generated millions of data points across arrayed and multiplexed assays, enriching the machine-learning models Ginkgo uses to explore and recommend new protein designs. (Read more about Ginkgo’s Enzyme Intelligence Services)
Ginkgo’s protein engineering experts determine which computational tools are best suited to the program’s goals. In parallel, our team is developing innovative generative AI models for protein design to tackle unique design challenges – work with us to take on challenges posed by pest resistance and herbicide tolerance. (Read more about our Ginkgo’s strategic partnership with Google Cloud)
Selecting for antibiotic activity with co-encapsulation and screening
The Challenge. Our partner uses a biocatalyst to perform a key transformation in their manufacturing process. They encountered a low synthesis to hydrolysis ratio (S:H) and formation of an undesirable byproduct. They partnered with Ginkgo to help increase enzyme activity, with the ultimate goal of reducing their manufacturing costs.
The Solution. Ginkgo created one enzyme variant library containing 1,100 unique strain candidates using molecular dynamics, Rosetta Design, and self-supervised machine learning models. In parallel, our team developed assay methods to simulate the partner’s process conditions for this enzyme. Using this assay, Ginkgo’s team performed 3,552 high- throughput strain tests across multiple time points. The results demonstrated a 6.5x improvement in activity and 5x improvement in desired product formation. Our activity improvement was confirmed by the partner; scale-up testing is in process.
03 GENE CONSTRUCT SYNTHESIS & DESIGN OPTIMIZATION
Ensure reliable and targeted delivery of multi-gene traits
Transformative traits often stack multiple genes that need to be expressed in specific plant tissues at specific times.
Ginkgo has deep experience with complex construct design and synthesis that deliver high quality constructs with short turnaround times. Our team tackles gene constructs that third party technology providers are unable to build by applying heuristics to reduce complexity (e.g., avoiding repetitive sequences) and delivers back-up designs to minimize risks of failed transformations.
In addition to optimizing protein performance, Ginkgo excels at codon optimization to make gene expression more reliable. Our codon optimization algorithm outperforms external vendors and literature state-of-the-art.
Ginkgo’s cell programming platform can also advance gene expression regulation in complex eukaryotic systems. Through promoter design and optimization, Ginkgo has demonstrated significant improvements in tightly regulated, inducible gene expression in mammalian systems.
Build constructs that other providers deem “not buildable”
By leveraging third party providers to make smaller constructs, Ginkgo’s Biofab can focus on “building the unbuildable”.
Optimize codons for reliable gene expression
Head-to-head with best alternatives, Ginkgo’s ML algorithm for codon optimization can deliver higher median and upper quartile expression.
Optimize promoters & other gene regulation elements for targeted gene expression
Leveraging multiplexed assays to test massive libraries of enhancer-promoter sequences, including natural and synthetic variants, under inducing and non-inducing conditions.
04 EFFICACY TESTING
Improve product stability and lower COGS ahead of commercialization
We can use several high to ultra-high throughput screening tools at various levels of confirmation to narrow down top performing variants. Our proprietary tools and deep experience in developing high throughput screening methods can be applied to disease control, insect control and herbicide tolerance.
Ultra high throughput screening with EncapS
Encapsulate variants into nanoliter reactors, generating microcolonies with reduced noise relative to single-cell approaches
Sorts and isolates high-performing microcolonies using fluorescence-based assays, including co-encapsulation with pathogens or insect proteins
Captures up to 1,000,000 cells per run in nanoliter-scale reactors.
READ MORE ABOUT OUR PARTNERSHIP WITH SYNGENTAHigh throughput in vitro assays
Test thousands of protein binding interactions using relevant activity assays, like receptor binding affinity.
Apply machine learning to further optimize and build libraries with higher likelihood of better performance
Plant Cell Pack technology
Accommodates any monocot or dicot plant species that is compatible with Agrobacterium transformation and cell culture
Up to 15,000 Agrobacterium-mediated transformation events in plant cell culture can be screened in a 384-well plate format
Protein expression hosts for sample generation
Leverage our proprietary strain assets with fully validated, high-performance fermentation processes, scalable from AMBR® 250 to 10,000L+
Pichia pastoris chassis strains demonstrating ~30 g/L intracellular protein production.
Aspergillus niger strains with demonstrated >20g/L secreted recombinant protein production
HOW WE PARTNER
An innovation partner with tech you can’t get anywhere else
Ginkgo Bioworks is a pure platform company. We’re dedicated to helping our partners discover and develop plant traits and ag biologicals rather than seeking to develop our own products in house.
We offer a variety of partnership models to meet your technical needs, whatever your stage of the R&D process. Our partnership philosophy is to align our commercial incentives so that we win when you win.
Our partners gain access to our industry-leading foundry automation infrastructure, our large libraries of partially and extensively characterized genetic assets and our world-class team of experts in agricultural biology.
Why work with Ginkgo?
- Take control of your R&D spending by replacing fixed costs with variable costs.
- Avoid costs and delays associated with building your own infrastructure.
- Gain access to our unique biological assets and industry-leading capabilities.
- Work with our team of dedicated experts across all stages of trait discovery and optimization
Ginkgo is ready to partner with you to enable the next generation of ag biologicals!