Solubility, stability, formulation, and manufacturability issues often appear after a molecule is chosen, forcing costly rework or restarts.
A molecule can be potent and selective yet still fail because it is hard to formulate, unstable, poorly soluble, or difficult to manufacture — and these developability problems often surface only after significant investment in a lead. Catching them earlier reshapes the candidate pool before commitments harden. AI approaches predict developability-related properties during design and selection so teams can weigh them alongside potency rather than discovering them downstream.
How AI helps assess developability earlier
Developability-prediction tools estimate the physicochemical and manufacturability properties that determine whether a molecule can become a practical medicine — solubility, chemical and physical stability, permeability, aggregation propensity for biologics, and related liabilities — from structure, before heavy investment. Folding these predictions into design and candidate selection lets teams treat developability as a first-class objective alongside potency and selectivity rather than a problem discovered later. The intent is earlier, better-informed trade-offs, not a guarantee. A model that flags likely poor solubility or an aggregation risk gives chemists and formulators a chance to steer the series or plan mitigation early. The predictions are hypotheses for experimental and formulation science to confirm; qualified scientists still run the studies and own developability decisions.
What to evaluate before buying developability-prediction AI
As with other property models, the decisive question is whether predictions hold for your modality and chemistry. Small-molecule and biologic developability involve different properties and different models, so confirm a tool was built for what you work on, ask how it performs on chemistry unlike its training data, and look for confidence or applicability-domain estimates rather than bare scores. Equally important is how developability fits your decision process: a prediction only helps if it arrives early enough to change design or selection, and if it can be weighed against potency and other objectives rather than viewed in isolation. Ask whether the tool integrates with your design workflow and whether it can be calibrated on your own formulation and stability data.
How teams typically get started
A practical entry point is retrospective benchmarking against molecules whose developability you already know — including candidates that ran into formulation or stability trouble downstream. Checking whether the tool would have flagged those problems earlier tests its relevance to your chemistry and modality before it shapes any live selection, and shows which properties to trust and which to treat cautiously.
AI Use Cases That Address This Problem
ADMET & Toxicity Prediction
Hit Identification & Lead Optimization
Frequently asked questions
What does developability mean in drug discovery?
Developability refers to the properties beyond target potency that determine whether a molecule can be made into a practical medicine — things like solubility, chemical and physical stability, permeability, and manufacturability. A potent compound with poor developability can stall or fail, which is why teams increasingly assess these properties early.
How does AI help catch developability problems early?
By predicting developability-relevant properties from a molecule's structure during design and selection, so teams can weigh them alongside potency rather than discovering them after a lead is chosen. The predictions guide where to focus and what to mitigate; experimental and formulation studies remain the source of truth.
Do developability models work the same for small molecules and biologics?
No. The relevant properties and the underlying models differ substantially between modalities — for example, aggregation and immunogenicity concerns are specific to biologics. Make sure any tool you evaluate was built and validated for the modality you actually work on, and check how it behaves on chemistry unlike its training data.
What should we ask a developability-prediction vendor?
Ask which properties and modalities the tool covers and how each was validated, how it performs on chemistry outside its training set, whether predictions include confidence estimates, whether you can calibrate on your own stability and formulation data, and whether output arrives early enough in your workflow to change design decisions.
Eli Lilly collaboration worth up to $670 million to discover novel therapies
Genesis Therapeutics (now Genesis Molecular AI) entered a strategic collaboration with Eli Lilly to discover novel therapies for up to five targets across a range of therapeutic areas using its Genesis Molecular AI drug-discovery platform, receiving $20 million upfront with a potential total value of up to $670 million.
Nimbus TYK2 program (TAK-279) acquired by Takeda for $4 billion upfront
A selective TYK2 inhibitor discovered using Schrödinger's physics-based computational platform (developed by Nimbus Therapeutics) was acquired by Takeda for $4 billion upfront, demonstrating commercial validation of a computationally designed clinical asset.
Eli Lilly AI and robotics drug-discovery collaboration
XtalPi and Eli Lilly announced a collaboration to apply XtalPi's AI and robotics-driven drug discovery platform to uncover first-in-class small-molecule therapeutics.
Cube Biotech collaboration to design novel Amylin Receptor agonists
Iktos and Cube Biotech announced a strategic collaboration to discover novel small-molecule agonists of the Amylin Receptor, pairing Iktos's AI-driven generative design platform with Cube Biotech's membrane protein production and purification technologies.
Pfizer collaboration expanded for next-generation molecular modeling
XtalPi announced an expansion of its research collaboration with Pfizer to build a next-generation molecular modeling platform for drug discovery, applying XtalPi's AI and robotics capabilities to improve the accuracy of computational predictions.
Pierre Fabre Laboratories oncology drug-discovery collaboration
Pierre Fabre Laboratories and Iktos announced an integrated drug-discovery collaboration to identify and develop novel small-molecule candidates in oncology, with Iktos applying its AI-driven generative design platform alongside Pierre Fabre's medicinal chemistry, biology, and preclinical development.