Julian Jude on Synthesizing a New Era in Functional Genomics

Disclosure: This post is sponsored by Twist Bioscience and reflects their views, opinions, and insights.

It’s a good time to be in functional genomics. The CRISPR toolkit is continually growing, providing researchers with the ability to manipulate genomes in ever more complex and large-scale ways. Multi-modal data can be analyzed using artificial intelligence (AI), enabling us to both design better functional screens and derive deeper meaning from them. Now more than ever, functional genomics studies can be carried out at scale and at speed. These era-defining advancements are only possible thanks to parallel advances in synthetic DNA manufacturing—an oft overlooked technology that is foundational to modern molecular biology. 

We recently sat down to chat with Julian Jude, Ph.D. about the evolution of synthetic DNA over the last decade and how it’s enabling this exciting era we live in today. 

Twist Bioscience has been producing synthetic DNA for more than a decade now. Over that time, the study of molecular biology has evolved considerably. How has Twist and the production of synthetic DNA grown with the field? 

Over the past decade, Twist Bioscience has grown well beyond the limits of being just a synthetic DNA company. We’ve developed into key players across the molecular biology space, with specific impact in NGS, biopharma, synthetic biology, and DNA data storage. This allows us to support everything from functional genomics and protein or genome engineering applications to antibody discovery and therapeutics development as well as storing digital data in DNA. And, we’ve been known to make some pretty great swag for the research community (we all love some good science socks). 

But through all this growth, our goal has remained the same: we create synthetic DNA tools to help our customers change the world for the better. A big part of fulfilling that goal is developing tools that give scientists the gift of time—time to discover, to innovate, and to do what they love doing. There’s a lot of tedious lab work (like cloning) that takes up precious time and distracts us from the things that excite us. When we empower researchers with tools that reduce this tedium and accelerate their projects, we all benefit. With our revolutionary DNA synthesis platform, Twist is uniquely positioned to create these tools.

We make these tools in the same way we always have, using our proprietary platform to produce highly accurate oligonucleotides. And we are very good at it. What’s changed is the boundaries of what’s possible in oligo synthesis, thanks in large part to our continuous process innovation. For example, our recent advances have set a new bar in the field by showing that it’s possible to routinely produce 500nt oligos (far exceeding previous length limits) through direct synthesis with high accuracy and scalability. 

As our abilities have grown, so too has our product offering. We’ve always said “friends don’t let friends clone,” because Twist can take care of this for you (we can ship clonal genes to you in just five days). Over time, we’ve continued to expand our abilities to meet the needs of a growing, evolving, and innovating market. We can now help users save time and ensure high accuracy in functional genomic screening projects by cloning entire oligo pools for them. Also, our ability to assemble huge, complex protein libraries by printing and recombining oligos with user-defined sequence variations allows researchers to design, screen, and develop better proteins for myriad applications. But, perhaps the best example of advances in the production and use of synthetic DNA is our latest release: Multiplexed Gene Fragments (MGFs), which have all the benefits of an oligo pool but the length of a gene fragment. This enables a whole new generation of high throughput experiments.

In short, we are constantly watching, listening, and responding to what scientists need.

What makes Multiplexed Gene Fragments so important? Is there a specific application area that will benefit most from them?

Multiplexed Gene Fragments are designed to solve a problem we’re all familiar with: You have an exciting hypothesis, but as you start designing the screen, your excitement turns to frustration as you realize that actually making the necessary synthetic DNA components is going to be a problem. Standard oligo lengths are too short to encode what you need, while the longer alternatives (generally known as DNA fragments) are not viable if you have high GC content or homopolymer repeats. Plus, the few options that are available to you don’t scale well, greatly limiting the throughput of your screen. So, what started as an exciting experimental design has quickly devolved into a meager compromise. 

We wanted to fix this, so Twist developed MGFs, which are purpose-built for high-throughput functional genomic and protein screening workflows. 

MGFs are double-stranded DNA fragments amplified directly off the chip that can be up to 500 bp long and are free of the limitations on sequence complexity (such as GC content) that normally affect gene synthesis. Already, we’ve synthesized complex MGFs for various laboratories, including some that encode for multiple gRNAs, long CRISPR arrays, and synthetic promoters with extreme GC content. Also 500 bases is the perfect size for engineering VHH antibodies (more on that in a moment). 

Beyond just helping with coding capacity, MGFs solve the scale issue as well. Pools start with 1,000 fragments and can scale to the size needed for your design. In most cases, with MGF the throughput of the screening workflow becomes the limiting factor, not the synthetic DNA. 

Looking towards the future, what are the most exciting potential applications of MGFs?

There is huge excitement about using machine learning (ML) and large language models (LLM) to help design or optimize genetic sequences, such as antibody CDR sequences and CRISPR arrays. New ML models emerge almost weekly, and the scientific community needs tools that help us keep pace. To really take advantage of AI and ML, we need the ability to both train design programs and test the many potentially valuable sequences they produce, regardless of complexity. 

MGFs are perfectly suited for this application, giving scientists the ability to directly synthesize the long (and potentially complex) designs for training and validation of models. In the past, this was severely limited by sequence degeneracy issues that forced researchers to abandon overly complex sequences and reduce the scale of their studies when possible. With MGFs, the breadth of sequences that can be tested grows significantly (at 500nt long, MGFs can encode roughly 20% of proteins listed on Uniprot and almost all known protein domains), greatly improving our ability to directly synthesize and screen these proteins without the need for codon optimization, enabling more accurately trained AI and ML models. 

We’re already seeing exciting applications of MGF-trained AI and ML algorithms in protein engineering and functional genomics. Unlike proteins which can be codon optimized, researchers studying the promoters, enhancers, and other “dark genome” elements are forced to test exact DNA sequences. The length and accuracy of MGFs empowers researchers to easily generate large libraries for MPRAs and bring the power of AI and ML to bear on functional genomics/epigenomics studies. 

Also, as mentioned, you can fit entire therapeutic proteins into 500 nucleotides. Recently, the research group of David Baker, one of this year’s Nobel Prize winners, used MGFs as a training dataset for an AI algorithm that could output new, strong VHH antibody binders to user defined epitopes. 

Overall, I think MGFs will grow to be a significant tool in molecular biology by effectively—and efficiently—bridging the gap between dry lab and wet lab work. And as our scientists continue to innovate and strive for a better planet, we’ll keep making products like MGFs to make it happen. 

Disclosure: This post is sponsored by Twist Bioscience and reflects their views, opinions, and insights.

About SBE

Established in 2004, the Society for Biological Engineering is a technological community for engineers and applied scientists integrating biology with engineering. Members of SBE come from a broad spectrum of industries and disciplines and share in SBE’s mission of realizing the benefits of bioprocessing, biomedical and biomolecular applications. Learn more about SBE