The paper introduces dynamic inference-time steering methods for de novo nanobody design, addressing the inefficiency of fixed-condition generative models. Two new budget allocation strategies are proposed: a joint Global Sequential Monte Carlo (SMC) approach that steers and allocates compute across multiple scaffold and epitope configurations simultaneously, and a decoupled Batched Successive Elimination (BaSE) method that adapts resource allocation during inference. Experiments show that these dynamic allocators boost the yield of designable candidates per GPU-hour by nearly fivefold while maintaining diversity across therapeutic targets and generative settings.
arXiv:2610.08367v1 Announce Type: new Abstract: Several discrete design tasks, such as molecular discovery, require diverse collections of useful candidates at low computational cost. High validity a...
The paper introduces MCTH (Monte Carlo Tree Hallucination), an inference-only framework that performs all‑atom biomolecular sequence‑structure co‑design by treating pretrained folding and inverse‑folding models as black‑box operators. MCTH uses Monte Carlo Tree Search to allocate a fixed inference budget across competing design trajectories, incorporating model confidence, uncertainty, and cross‑expert consensus. Experiments across protein‑RNA, protein‑DNA, protein‑protein, and protein‑ligand design show that adaptive search outperforms simpler sampling strategies, and evaluations with AlphaFold3 and Chai‑1 demonstrate transferability beyond the search‑time oracle.
arXiv:2610.07560v1 Announce Type: new Abstract: Goal-directed molecular design has advanced rapidly, yet a substantial proportion of designed molecules remain difficult to synthesize in practice, lim...
arXiv:2607. 02834v1 Announce Type: new Abstract: Molecular optimization often starts from a pretrained generative model that captures a broad prior over valid molecular structures.
arXiv:2609.00189v1 Announce Type: new Abstract: Goal-directed optimization is essential for steering molecular generators to propose candidates with desired properties. However, it is often implement...
arXiv:2606. 01220v1 Announce Type: cross Abstract: Generating molecules that simultaneously satisfy drug-like properties and conform to the 3D structure of a target protein is a core challenge in structure-based drug design (SBDD).
arXiv:2509. 26405v2 Announce Type: replace Abstract: We introduce InVirtuoGen, a discrete flow generative model for fragmented SMILES for de novo and fragment-constrained generation, and target-property/lead optimization of small molecules.
arXiv:2607. 12488v1 Announce Type: new Abstract: Molecular optimization in drug discovery, materials design, and catalysis requires searching vast chemical spaces under tight evaluation budgets, since high-fidelity oracles and experimental measurements are costly.
The paper proposes a new framework called ensemble-conditioned guidance that reframes molecular design as an optimisation over both the modes and properties of a molecule’s conformational ensemble. It allows 3D generative models to be conditioned simultaneously on multiple axes—such as shapes, pharmacophore profiles, or protein pockets—by adaptively combining vector fields from each condition. The authors introduce adaptive symmetry learning for composable conditions across reference frames, extend the framework to support flexible-size generation, and demonstrate its effectiveness on new benchmarks and practical drug‑discovery tasks, showing improved outcomes when conditioning on additional states compared to single‑state approaches.
arXiv:2606. 26657v1 Announce Type: new Abstract: Identifying high-utility candidates from massive discrete spaces under expensive evaluations is a recurring challenge across the sciences, with structure-based drug discovery as a prominent example.
The paper introduces a new Bayesian optimization approach tailored for generative models used in de novo discovery pipelines. By employing a linear surrogate model constrained to a spherical domain—where high‑dimensional latent vectors naturally concentrate—the authors derive nearly closed‑form solutions for both surrogate modeling and acquisition, achieving at least a 100‑fold speedup over existing methods. This acceleration enables Bayesian optimization to be used as a practical drop‑in component in pipelines that previously found it too slow to consider.
NEAT-POCKET is a pocket‑conditioned extension of the autoregressive NEAT model that generates 3D molecules atom by atom within protein binding pockets, maintaining atom permutation invariance and explicitly modeling hydrogen atoms. It outperforms existing baselines on the CrossDocked and SPINDR datasets, achieving competitive structure‑based generation performance while sampling significantly faster. The model also supports pocket‑conditioned fragment completion, a capability directly useful for lead optimization and scaffold elaboration in drug design.