In the traditional landscape of small-molecule drug discovery, the timeline is often dictated by a rigid, linear progression: hit identification, hit-to-lead, and finally, lead optimization. For decades, the comprehensive profiling of Absorption, Distribution, Metabolism, and Excretion (ADME) properties has been reserved for the latter stages of this process. It is a costly, time-intensive phase where programs funnel significant capital into optimizing compounds that have already cleared initial biological screens.
A new collaborative venture, however, aims to collapse this traditional funnel. Ginkgo Datapoints, Tangible Scientific, and Inductive Bio have officially launched "ADME-One," a high-throughput, integrated service platform designed to pull critical pharmacokinetic (PK) projections from the tail end of discovery directly into the hit identification phase. By merging automated laboratory science, sophisticated compound management, and machine learning (ML), the partners hope to ensure that drug developers can identify "dead-end" molecules before they consume months of precious synthesis time and research budget.
The Structural Mechanics of the ADME-One Platform
The ADME-One platform functions as a tightly knit ecosystem designed to eliminate the friction points typical of outsourced drug discovery. At its core, the platform synthesizes three distinct technological pillars:
- Ginkgo Datapoints (Automated Tier 1 Assays): Operating from its state-of-the-art laboratory in Boston, Ginkgo executes five critical Tier 1 assays: microsomal stability, cell permeability, kinetic solubility, CYP inhibition, and plasma protein binding. By automating these processes, the platform provides the high-quality empirical data necessary for modeling.
- Tangible Scientific (Logistics and Sample Custody): Managing the physical flow of compounds is often an overlooked bottleneck. Tangible Scientific assumes full custody of physical samples, managing intake, precise plating, and real-time tracking, ensuring that the transition from a digital design to a physical test is seamless.
- Inductive Bio (Human PK Projection): The "brain" of the operation is Inductive Bio’s Compass platform. It ingests the raw data from Ginkgo’s assays and synthesizes them into actionable human PK projections, allowing medicinal chemists to visualize the likely performance of a compound in a clinical setting long before it ever reaches an animal model.
A Shift in Chronology: From Late-Stage Validation to Early-Stage Insight
Historically, the decision to invest heavily in a compound’s ADME profile was deferred to minimize the cost of "failed" molecules. However, this creates a dangerous paradox: by the time a team discovers a compound has poor metabolic stability or high risk for drug-induced liver injury (DILI), they may have already committed thousands of dollars in synthesis and weeks of specialized labor.
"We asked: Could we pull together all the assays needed to get your first projection of human PK at a price point where you’d now be doing this on most, if not all, of the compounds coming through?" says Alex Taylor, Ph.D., head of medicinal chemistry at Inductive Bio.
By making this a standard "Tier 1" activity, the partnership effectively shifts the discovery milestone. Instead of asking if a lead is "potency-ready" at month six, the team asks if it is "dose-ready" at week one.
Supporting Data: Why "Dose" is the Ultimate North Star
The urgency for earlier PK projections is driven by a growing recognition of the role "human dose" plays in clinical success and safety. Medicinal chemists have long understood that while potency is essential, a molecule that requires a massive daily dose to achieve its effect is inherently more difficult to move through clinical trials.
The Link Between Dose and Toxicity
Data from the FDA and various registry studies underscore the risk associated with high-dose drugs. For example, research published in Hepatology has highlighted the "rule-of-two," noting that high daily doses—particularly when paired with high lipophilicity—are strongly correlated with a risk of drug-induced liver injury. Further registry studies have indicated that drugs requiring doses of 50 mg or more per day face statistically higher rates of liver failure and associated medical complications compared to those dosed at 10 mg or below.
Simplifying the Therapeutic Profile
Beyond toxicity, there is a practical imperative for lower-dose compounds: patient adherence. Simpler dosing regimens are notoriously more effective in real-world patient populations. However, calculating a projected human dose has historically been an elusive, data-heavy task. ADME-One aims to bridge this gap by integrating the ADME variables required to forecast the human dose at the earliest possible stage of development.
Addressing the "Irreconcilable" Profile
The platform also addresses the nuances of drug design where simple, binary "pass/fail" metrics fail to capture the full picture. Dr. Taylor points to the evolution of triazole antifungals as a classic example. Fluconazole, which is polar and cleared renally, stands in stark contrast to itraconazole, which is highly lipophilic and cleared via hepatic metabolism. On a standard assay scorecard, one might assume both would fail conventional screening criteria. Yet, both became essential medicines. ADME-One seeks to empower chemists to recognize this "balance of properties" earlier, preventing them from discarding potentially viable scaffolds due to narrow, inflexible criteria.
Official Responses and Strategic Implications
The launch of ADME-One is as much a response to current macroeconomic pressures as it is to scientific necessity. In an era where biotech capital is constrained, the mandate to be "lean and cost-conscious" is universal.
The Reshoring of Preclinical Work
The platform is explicitly designed to compete with, and ultimately beat, the pricing structures of offshore Contract Research Organizations (CROs). This is a strategic move, particularly as U.S. and European developers face increased scrutiny regarding data sovereignty and the implications of the BIOSECURE Act. By keeping the workflow entirely within the United States, ADME-One provides a layer of geopolitical security that is becoming increasingly important to boards and investors.
The Consortium Model: Security and Innovation
One of the most complex challenges for Inductive Bio was addressing the "data paradox"—how to improve global machine learning models using proprietary customer data without compromising the confidentiality of any individual firm’s chemical library.
Dr. Taylor explains that Inductive Bio maintains a strict legal and technical firewall. The consortium model functions as a pooled database where the collective intelligence of all partners is used to sharpen the global model. When a new customer uses the platform, their data is used to fine-tune a "local" model on top of the global one, which has consistently shown to yield significant performance gains. Furthermore, the engineering team has implemented robust security measures designed to prevent "reverse-engineering" of the underlying chemistry, ensuring that no participant can infer the structure of another partner’s molecules from the pooled insights.
The Virtuous Cycle of Discovery
The long-term vision for ADME-One is the creation of a "virtuous cycle." In this model, the design phase is continuously informed by the previous round of empirical data.
- Design: Chemists use the Inductive Bio platform to visualize predicted ADME and PK curves.
- Test: Selected molecules are synthesized and run through the Ginkgo/Tangible ADME-One workflow.
- Learn: Experimental results are fed back into the system. This verifies the predictive accuracy of the model and provides fresh, high-quality data to the consortium.
As this cycle repeats, the global models become increasingly precise. "It’s a flywheel," Dr. Taylor notes. "The more chemical matter we feed the consortium, the better the predictions become for everyone."
The Human Element: Science vs. Engineering
Despite the sophisticated automation and ML integration, the team is careful to frame their platform as an aid to—not a replacement for—human judgment.
"Drug discovery is science at the end of the day, and science is not engineering," Taylor emphasizes. "You can give your best guess of what a compound is going to do, but at the end of the day, you need to reduce it to practice."
The ultimate utility of the ADME-One platform lies in its ability to optimize the decision-making process. By providing faster, cheaper, and more accurate data on the "front end," the platform allows researchers to decide which compounds are truly worth the high cost and time of synthesis. In a field where the "cost of failure" is astronomical, this shift in the discovery timeline could fundamentally redefine the success rate of small-molecule development programs across the industry.
As the partnership matures, the impact of this "shift-left" strategy will likely be measured in the reduced duration of preclinical programs and, eventually, a higher throughput of safer, more efficacious drugs reaching the clinic.
About the Author
