Exscientia is one the best kept secrets in drug development and investing hiding in plain sight.

People swooned over Schrodinger. Rushed into Recursion Pharmaceuticals. They... Well, they still mostly ignore Relay Therapeutics (kidding, it has the highest market cap now). But if that's the bar for ignorance, then Exscientia is virtually nonexistent in the eyes of the market.

Similar to the peers mentioned above, the U.K. company is a technology-enabled drug developer utilizing large datasets (i.e. artificial intelligence) to increase the efficiency of drug discovery. Unlike the peers mentioned above, Exscientia has the most experience and longest track record of execution.

The company owns the first, second, and third ever AI-designed drug candidates to enter clinical trials. It also wields the first AI system demonstrated to improve clinical outcomes in cancer by optimizing treatments for patients. The last part is an important distinction.

Whereas most peers are eager to improve the probability of success (POS) of drug development by designing better drug candidates before entering clinical trials, Exscientia has proven it can also leverage its patient-first datasets to match an individual's cancer to the best treatment option among all available therapies. That's potentially pretty valuable. This isn't slapping wishy-washy promises of the wonders of AI and robotics and [insert trendy techbro thing here] on an investor deck for the sake of innovation porn. This is improving real-world outcomes for real humans.

Talk Nerdy to Me

The most important metric for understanding investments in drug developers is the probability of success (POS).

• At a high level the POS is the percent chance that a drug candidate in clinical trials will reach the market and generate revenue.
• Drilling one level deeper, the POS changes as a drug candidate progresses through clinical trials. An asset in a phase 1 clinical trial will have a lower POS than an asset in a phase 3 clinical trial.
• Digging deeper still, the POS is often applied to the risk-adjusted net present value (rNPV) model attempting to understand the level of revenue, profit, and cash flow an experimental asset might generate if it reaches market. An asset in a phase 3 clinical trial contributes more to a company's valuation because investors can pencil in a higher percentage (the POS) of a rNPV model. An approved drug product has a POS of 100%, which means all of a rNPV can be factored in (of course, the model can still be wrong).

Investors often question how fair value can be calculated for an unprofitable or even pre-revenue company. This is especially true for precommercial drug developers. The answer: It all comes down to POS and rNPVs.

Solt DB Invest uses relatively granular and nuanced POS estimates and rNPV models using our bottom-up approach to biotech, synthetic biology, and living tech stocks.

• For example, an oncology drug candidate in a phase 1 clinical trial has a roughly 5.3% chance of reaching market, which rises to 43.9% once a phase 3 clinical trial begins.
• An ophthalmology drug candidate in a phase 1 clinical trial has a roughly 11.9% chance, which rises to 46.7% chance once a phase 3 clinical trial begins.

Those are averages from the last decade, which can then be tweaked based on safety signals from the drug class, selectivity of the specific drug candidate, and so on.

However, as the examples in the bullet points above make clear, most drug candidates fail in phase 2 clinical trials regardless of therapeutic area or therapeutic modality. That really sucks. That's a really inefficient way to develop new drugs. Why? It takes years and perhaps $100 million to reach a mid-stage study. Ideally, companies want to develop the most accurate POS models for drug candidates before they enter clinical trials, not while they're being studied in phase 2 clinical trials.

This is the central promise of technology-enabled drug developers such as Relay Therapeutics and Exscientia. If these companies can refine their unique technology platforms for drug discovery and drug design, then they should be able to greatly reduce failure rates (and therefore increase their POS) by bringing better drug candidates into clinical trials from the start.

A Unique Tech-Enabled Approach

What sets Exscientia apart from peers in the loosely-defined category is the ability to apply its AI-models at any stage of development, not just drug discovery and drug design. Additionally, the company can apply its platform to both small-molecule drugs and more complex biologic drugs.

These advantages are driven by the company's unique approach. Exscientia has integrated single-cell analysis of human tissues from real patients into its platform from discovery through clinical development. That provides cleaner and clearer insights into how drug candidates react with human and tumor biology compared to, say, human tissues grown from standard cell lines or animal cell cultures. After all, human tissues from real patients are more representative of the environment a drug candidate will encounter inside the body.

This iterative approach can be used at either end of the drug development spectrum:

• Discovery: By understanding human tissues from real patients, Exscientia has more insights into what molecular signals are important for a specific disease. In certain cancers it's important to inhibit both CDK4 and CDK6. However, dual inhibition often leads to resistance mutations that make CDK4/6 drugs ineffective. The company used its unique approach to discover the role CDK7 plays in determining outcomes for individuals with CDK4/6 resistance mutations. It designed a drug candidate capable of selectively inhibiting CDK7 at low doses, that's rapidly absorbed, and may have a tolerable safety profile. It'll need to thread a needle of pharmacodynamics and pharmacokinetics metrics, but if correct, then the asset could potentially treat breast, lung, and other cancers.
• Therapy Optimization:The EXALT-1 study used the company's platform to match cancer patients with optimal treatments for their individual cancers. By taking tissue samples from each individual, Exscientia could test over 100 approved drug compounds on the tumor across a range of metrics and determine the likely most effective drug. It's essentially as if the technology platform is a massive diagnostic product that could be used across multiple cancer types.
• 54% of patients demonstrated a clinical benefit of more than 1.3-fold enhanced progression-free survival compared to previous therapy
• 40% of patients experienced exceptional responses lasting at least three times longer than expected for their respective disease

This is a truly patient-centric approach. It's also catching the attention of the global industry.

Exscientia counts over 29 programs in early discovery stages or later, spanning relationships with eight unique customers, and complemented by a wholly-owned pipeline. That's much larger than the pipeline at Relay Therapeutics. The U.K. based drug developer could collect over $3.5 billion Across all programs and partnerships.

Intriguingly, the company has done a better job managing operating expenses than Recursion Pharmaceuticals, a peer with a similarly large pipeline. In the first nine months of 2022, Exscientia reported an operating cash outflow of only $15 million thanks to healthy payments from collaborators. It could collect over $3.5 billion in precommercial milestones (and over $6.5 billion total) from its stable of partners.  The business ended September with $625 million in cash.

Forecast & Modeling Insights

To be provided in the next article.

Margin of Safety & Allocation

(No change.)

Exscientia is considered a Growth (Speculative) position. The current margin of safety range for the company is below:

• Current Price (market close January 31):  $7.04 per share
• Likely Undervalued:          <$7.27 per share
• Midpoint:                           $9.57 per share
• Likely Overvalued:            >$11.86 per share
• Allocation Range:             Up to 5%

Exscientia has an estimated 122.878 million shares outstanding as of January 30, 2023 (unconfirmed). The margin of safety range above does not account for dilution, but will be updated following the fourth-quarter 2022 conference call.

Further Reading

• October 2021 press release discussing results from the EXALT-1 clinical trial

‍