As it builds a pipeline of artificial intelligence (AI)-designed treatments, Absci has reached out to its preferred supplier of DNA for its latest collaboration—one of three partnerships that the AI drug developer has either launched or advanced in recent weeks.
Twist Bioscience says the potential for developing an AI-based drug is great enough to depart from its usual business model as it joins Absci in designing a novel therapeutic using generative AI.
The companies have agreed to apply Twist’s silicon-based high-throughput gene synthesis platform and Absci’s Integrated Drug Creation™ platform toward their combined drug discovery effort, which is designed to accelerate the design and preclinical development of an antibody therapeutic for an undisclosed set of diseases.
The undisclosed disease set is expected to be one of Absci’s three therapeutic areas—inflammatory bowel disease (IBD), dermatology, and immuno-oncology, Absci founder and CEO Sean McClain told GEN Edge.
“We do see this falling within those broader I&I and oncology therapeutic areas of focus,” McClain said.
Emily M. Leproust, PhD, Twist’s CEO and co-founder, told GEN Edge the collaboration with Absci marks a departure from her company’s traditional fee-for-service business model.
“I call this partnership the cherry on top of the cake,” Leproust said, “because our usual measure of success for us is revenue ramp, gross margin improvement, getting to profitability. Our day-in, day-out focus of what we do is usually to sell a service or sell a product: Someone gives us a sequence, we give them a tube, they pay us.”
For the Absci-Twist collaboration, she explained, both companies have agreed to fund their own costs and share the resulting asset.
Creating value
“Hopefully it will create an asset that will be valuable, and we’ll share in that value creation,” Leproust said. “It’s an opportunity for us to participate in the value creation of a new asset, as well as show to the world how the new technology that we’ve developed is particularly suited for AI developed antibody discovery.”
The companies will test and validate antibody candidates designed using Absci’s AI de novo design capabilities, combining that technology with Twist’s silicon-based synthesis platform. Twist’s platform includes its Multiplexed Gene Fragments, double stranded gene fragment sequences of up to 500 base pairs in length with no limit on sequence number, delivered as a pool instead of one fragment per well matrix. Twist says the pooled format of fragments enables high-throughput screening applications such as prime editing, ultra-complex CRISPR-based functional screening, peptide and protein engineering.
“With 500 base pairs, you can synthesize a full heavy chain or a full light chain of an antibody. In the past, when you thought antibody, you were always thinking, I either have to restrict myself to a portion of a heavy chain, light chain, or I have to assemble. The first part is not ideal, and the second part takes time,” Leproust said.
“Now that we have those multiplexed gene fragments, because we can do the full heavy chain or full light chain, you get the full antibody and you get speed. That accelerates the design-build-test cycle,” she added.
Customers who furnish Twist with an antibody sequence can receive an antibody tube completed in as little as 13 days, compared with as much as six weeks using the previous fragments of up to 300 base pairs long. Twist says its platform enables it to produce high quality gene fragments starting at starting at 7 cents per base pair (bp) and next-generation sequencing (NGS)-verified sequence perfect clonal genes starting at 9 cents per bp.
Twist launched its Multiplexed Gene Fragments in May. Three months later, it expanded its DNA synthesis offering by launching Gene Fragments with increased lengths ranging from 1.8 kb to 5.0 kb.
While the value of the collaboration has not been disclosed, Twist says the collaboration is the company’s first with an AI drug developer.
“A little bit of a risk”
“We’re both taking a little bit of a risk,” Leproust added. “Frankly, this collaboration is not something we can do too many of because our business model is to ramp revenue and be paid. But we’re very happy to do it with Absci because I think they have a great platform.”
“If it works great, maybe we’ll do more.”
For Absci, McClain said, the partnership is the first with a partner that does not involve a biopharma company or a research institution: “This is a unique partnership for us, and I think something from which we see a lot of potential for value creation.”
“One of the things that we’re seeing is that there are a lot of exciting targets that can be prosecuted on,” McClain explained. “We don’t have the ability to prosecute on all of these ourselves. And so being able to co-develop these allows us to essentially prosecute on more assets ourselves than we could do on our own
As the partnership progresses, Absci and Twist plan to seek a partner for human clinical development and commercialization.
“Once we have some of that initial preclinical data, especially around a drug candidate, I think that’ll be the time we take this data to a big biotech or large pharma,” McClain said.
Also in recent weeks, Absci disclosed that it had successfully completed the first milestone in its nearly year-old drug discovery collaboration with AstraZeneca—namely, successfully delivering AI de novo designed antibody sequences to the pharma giant.
“They have acknowledged the successful completion of the milestone and have elected to move forward” with advancing the antibody sequences into AI-lead optimization, McClain said.
As a result of achieving that milestone, Absci received an undisclosed payment from AstraZeneca. The companies “The agreement was structured to have technical milestone payments in there, along with the other traditional clinical and development milestones as well,” McClain added.
The companies agreed in December 2023 to partner on accelerating the discovery of a potential new cancer treatment candidate, with AstraZeneca agreeing to pay Absci an upfront payment, R&D funding, payments tied to achieving milestones, plus royalties on product sales—all of which were undisclosed.
The collaboration was designed to combine Absci’s platform with AstraZeneca’s oncology expertise. Oncology is one of AstraZeneca’s five therapeutic areas of focus, along with cardiovascular, renal, and metabolism; respiratory and immunology; vaccines and immune therapies; and rare diseases.
Work on the partnership started early this year.
Generating binders
“Within six months, we were able to use our de novo AI model to generate binders to this oncology target of interest. And this was designing all six CDRs [antibody complementarity-determining regions] from scratch; there were no known binders,” McClain explained.
“We’ve obviously applied this to our own internal pipeline, but applying it successfully to a partner program is something we see as a huge achievement as well as a very strong validation of the model and the platform,” McClain added. “We’re planning on moving forward into the next stage of the partnership. We’re really excited.”
Numerous leads toward the target have been generated, from which only one candidate will be selected to advance toward the clinic. The companies have yet to finalize a timeframe for clinical study, McClain said.
Absci’s all-preclinical pipeline is led by ABS-101, which is designed to treat IBD by targeting tumor necrosis factor-like cytokine 1A (TL1A). The company remains on track for an early 2025 launch of a first-in-human Phase I trial of ABS-101, McClain said.
Also in Absci’s pipeline:
- ABS-201, which is being developed for an undisclosed dermatological indication that according to the company has significant unmet need.
- ABS-301, a fully human antibody that is designed to bind to a novel target discovered through Absci’s Reverse Immunology platform.
In August, Absci broadened its cancer pipeline by agreeing to partner with Memorial Sloan Kettering Cancer Center (MSK) on using generative AI to discover and develop up to six cancer therapies, through a collaboration intended to marry Absci’s platform with MSK’s research and clinical expertise in oncology.
Absci and MSK have not disclosed the value of their collaboration or the types of cancer for which they plan to develop new treatments.
“As a leading cancer research institute, MSK has a lot of exciting new novel oncology targets to prosecute. We’re going to take that novel biology and cancer expertise along with our generative AI models and be able to develop some potential first-in-class therapeutics together,” McClain said. “It’s another really strong validation of the platform.”
Absci is the first AI-based drug developer with which MSK has partnered. The partners have offered no public updates since announcing their collaboration, though McClain said that will change when Absci holds its R&D Day on December 12.
MSK and Twist are two of four planned partnerships Absci planned to announce during the second half of this year. Absci still plans to reveal the other two by year’s end.
