Ginkgo Datapoints, Tangible Scientific, and Inductive Bio Launch ADME-One, a Packaged ADME Service for Early Discovery

Ginkgo Bioworks' (NYSE: DNA) Datapoints arm, Tangible Scientific, and Inductive Bio today launched ADME-One, a single-SKU service that combines a Tier 1 ADME panel, AI-driven human pharmacokinetic projection, and compound management. It is the first product to come out of the three-way partnership the companies announced in August 2025, which committed them to a lab-in-the-loop workflow in which experimental ADME data flows back into Inductive's models for retraining.

ADME Testing During Hit Identification

Tier 1 ADME profiling — microsomal stability, cell permeability, kinetic solubility, CYP inhibition, and plasma protein binding — has historically been priced and scheduled for lead optimization. That means teams often pay to confirm liabilities in molecules they have already committed months of work to. The partners argue ADME-One is cheap and fast enough to run during hit identification, so that experimental ADME data feeds into series-level decisions from the start.

Within the initiative,

• Ginkgo's Boston facility runs the five assays on its automated lab;
• Inductive Bio's Compass platform takes the resulting endpoints and outputs a human PK projection, collapsing the panel into a single optimization endpoint that teams can rank compounds against;
• Tangible handles compound intake, plating, and tracking from its Natick, MA facility.

The commercial framing rests on two arguments: ADME-One is entirely U.S.-based and priced, the partners say, below offshore CROs — relevant to sponsors reshoring preclinical work in response to the BIOSECURE Act; and Tangible's involvement is meant to close the operational gap between AI-driven compound design and the wet-lab work that has to validate it.

Earlier Validation with ADMET Models

Inductive's modeling work has performed well in two recent public benchmarks. Its Beacon models placed first in the OpenADMET–ExpansionRx blind challenge against more than 370 submissions from pharma, biotech, academia, and AI groups, following a first-place finish in the 2025 ASAP–Polaris–OpenADMET antiviral drug discovery challenge.

The ExpansionRx challenge asked participants to predict nine ADMET endpoints from a real drug discovery campaign targeting RNA-mediated diseases. The OpenADMET organizers' own follow-up analysis showed Inductive's winning model substantially outperformed zero-shot public tools like ADMETlab 3.0 and ADMET-AI when trained on the campaign data.

Ginkgo also said it has hired Jonathan Grob, previously at Novartis and Valo Health, as VP of Small Molecules to lead ADME-One product strategy. The three companies will share validation data and case studies at the New England Drug Metabolism Discussion Group meeting in Boston on May 27.

Topic: AI in Bio