How generative AI could improve drug discovery


Absci is a generative artificial intelligence (AI) drug creation company with goals of bringing better biologics to patients more quickly. Here, DDW Editor Reece Armstrong speaks to Sean McClain, the Founder and CEO of Absci about how generative AI is being used within drug discovery and the company’s work in the antibodies sector.

RA: Could you explain the difference between drug discovery’s usual approach to AI and the use of generative AI within the sector?

SM: Conventional drug discovery depends on using protein-protein interaction libraries to try and find existing potential drug candidates. Not only is that like searching for a needle in a haystack, but the needle you find might not even work as well as you want it to. At Absci, we take a different approach – we create the needles. We combine generative AI and our wet lab technology to create new large-molecule medicines in silico that don’t exist in nature. And we can do this because we have the massive, high-quality biological data that generative AI needs to create better drugs.

RA: Will the use of this type of AI help generate therapies that target previously ‘undruggable’ diseases?

SM: The term ‘undruggable’ historically has described diseases that drugs couldn’t help. Instead of calling them ‘undruggable’ diseases, I think of them as ‘not drugged yet’. These diseases often present challenging targets where we’re creating new potential therapies for the first time. The good news is that, thanks to AI and synbio, we’re making real progress in developing better biologic drugs, faster. If we can increase clinical success rates and bring large-molecule medicines to market faster, it will increase the return on investment (ROI) for these diseases. Eventually, the industry will reach a tipping point where it becomes more and more feasible to pursue smaller and smaller patient communities. When that happens, it will signal a true beginning of the era of personalised medicine, and we’ll be one step closer to the vision of breakthrough therapies at the click of a button.

RA: Through its platform, Absci has begun generating antibody designs, but does it have it plans to advance any of these concepts through clinical trials?

SM: Yes, we are advancing partnership programmes as well as our own internal drug candidate pipeline of de novo antibodies designed in silico.

RA: How will the use of this type of AI help researchers in drug discovery labs?

SM: Generative AI is a drug discovery game changer. Our wet lab is capable of testing and validating nearly 3 million unique AI-generated designs each week. It would take others years to generate that kind of data. Our proprietary ACE assay technology and wet-lab capabilities allows us to analyse high-quality protein- protein interactions, and refine our designs even further. This feedback loop is an invaluable component of bringing better antibody candidates to the clinic quicker.

RA: How impactful can this tech be in terms of reducing time to market and more wider cost savings?

SM: It currently takes about 10 years and costs well over $1 billion to get one new drug to market—with only about a 5% success rate. The cost of drugs reflects the high risks of development – we pay not only for the drugs we’re taking but also for the other ~95% of drugs that never made it to market.

Our generative AI approach simultaneously optimises drug candidates for affinity, naturalness, and manufacturability from the very start, which we believe will increase the probability of success in the clinic. This has the potential to reduce the cost to drug makers and patients alike by reducing the risk of failure in the early stages.

In addition to potentially increasing clinical rates of success, we believe our approach to drug creation can reduce preclinical development times of four to six years by over 50%. If you have a chance of being significantly faster in today’s highly competitive drug discovery environment, that puts you in a completely different competitive situation.

RA: Can generative AI improve the chances of success in bringing a therapy to market?

SM: Yes, and our de novo antibody breakthrough provides the pharmaceutical market with an alternative to the traditional drug discovery process. Now companies can design antibodies with binding affinities to increase their chances of success in the clinic. If this breakthrough only increases the chance of success in the clinic from, for example, 5% to 10%, that would have huge financial and societal implications.

RA: Could you go into some detail regarding the de novo antibodies Absci has created?

SM: Absci used zero-shot generative AI to design the heavy chain complementarity determining region 3 (HCDR3) from scratch. HCDR3 is a critical region for antibodies to bind to their targets and enable their therapeutic potential.

The key to this accomplishment was zero- shot generative AI, a method that involves designing antibodies to bind to specific targets without using any training data of antibodies known to bind to those specific targets. Our zero-shot model generated antibody designs that were unlike those found in existing antibody databases. The AI-designed antibodies generated directly from our models were tested and functionally validated in the wet lab – without the slow and costly step of further optimising the in silico designs in the lab. We further validated antibodies for HER2 and multiple additional targets.

We think this represents a major industry breakthrough on the path to fully de novo antibody design and our vision to deliver breakthrough therapeutics at the click of a button, for everyone.

RA: Absci’s focus is currently on antibodies but does this platform extend to other disease areas?

SM: Absci is focused on antibodies because of their versatility and specificity. Though the pharma industry has made tremendous progress on antibody therapeutics in recent decades, we have barely begun to realise the potential of this class of drugs. In the future, our vision is to use our platform to deliver breakthrough therapies at the click of a button for everyone. If we’re successful, it will represent the start of the era of personalised, precision medicine.

DDW Volume 24 – Issue 2, Spring 2023


Sean McClainSean McClain is the Founder and CEO of Absci, a generative AI drug creation company on a mission to create better biologics for patients, faster. Under his leadership, Absci has scaled to over 200 employees, raised over $450M in capital, established a growing number of industry-leading partnerships, and built an Integrated Drug Creation Platform that merges bleeding-edge generative AI models with proprietary high-throughput wet-lab data.

Related Articles

Join FREE today and become a member
of Drug Discovery World

Membership includes:

  • Full access to the website including free and gated premium content in news, articles, business, regulatory, cancer research, intelligence and more.
  • Unlimited App access: current and archived digital issues of DDW magazine with search functionality, special in App only content and links to the latest industry news and information.
  • Weekly e-newsletter, a round-up of the most interesting and pertinent industry news and developments.
  • Whitepapers, eBooks and information from trusted third parties.
Join For Free