arXiv:2606. 00555v1 Announce Type: new Abstract: Structure-based drug design increasingly employs LLM agents to iteratively refine ligands against a target pocket, yet a viable ligand must satisfy two often-conflicting objectives -- binding affinity and druggability -- which single optimization steps rarely improve together.
arXiv:2608. 11483v1 Announce Type: new Abstract: Hit-to-lead optimization requires iterative design of hit analogs across competing potency, selectivity, physicochemical, pharmacokinetic, safety, and synthetic constraints.
SpecOpt is a new molecular design task that optimizes the binding specificity of existing drugs by making constrained structural modifications. The method uses an agentic framework that docks a compound against its intended target and known off‑targets, compares residue‑aware atom‑protein contacts, and feeds the differential interactions to a large language model to propose changes. On a benchmark of 915 compounds, SpecOpt increased the target‑off‑target binding gap for 84.8% of cases while preserving drug‑like properties and structural similarity.
SurfSpec is a new framework for lead optimization that improves drug specificity without needing off‑target structures. It works by measuring and reducing the geometric mismatch between a ligand and its target pocket, using the triangle inequality to infer a lower bound on mismatch to off‑target pockets. In tests on the CrossDocked2020 dataset, SurfSpec lowers geometric mismatch and achieves higher empirical specificity while still improving target affinity.
SurfSpec is a lead‑optimization framework that improves drug specificity without needing off‑target structures. By measuring and reducing the geometric mismatch between a ligand and its target pocket, SurfSpec provides a conservative lower bound on specificity against geometrically separated off‑target pockets. The method iteratively grows ligands toward under‑occupied target surface patches, alternating between linker generation and refinement, and demonstrates superior empirical specificity on the CrossDocked2020 test set while maintaining competitive target affinity.
arXiv:2608. 19906v1 Announce Type: new Abstract: Accurately ranking active ligands for a target protein pocket from massive chemical libraries remains a central challenge in virtual screening.
pCoMole is a new framework that uses discrete flow matching to guide molecule editing toward user-specified multi-objective preferences while enforcing hard feasibility constraints. It introduces a feasibility-gated terminal distribution with an augmented Tchebycheff utility and implements the preference tilt via a Doob‑h transform, approximated with short Monte Carlo rollouts for efficiency. The method is validated on tasks such as shrinking GFP, shortening Cas9 orthologs, and compressing peptide binders, with wet‑lab tests showing that designed eGFP variants retain fluorescence after extensive edits.
ProbeMatchDTI is a new framework for drug‑target interaction prediction that uses probe‑driven pattern matching to preserve weak biochemical signals. It introduces IterProbe, which retains contextual states across refinement depths and selects them with learnable probes, and BindingProbe, which models drug‑protein complementarity at both local and whole‑pair levels. Experiments show that ProbeMatchDTI outperforms existing methods, improving AUC‑ROC by 2.0% on BindingDB and 0.5% on DrugBank, and its predictions can be integrated into downstream drug‑discovery workflows.
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.
ProbeMatchDTI introduces a probe-driven framework for drug‑target interaction prediction that preserves weak biochemical signals by using IterProbe to retain contextual states and BindingProbe to model cross‑entity complementarity at multiple scales. The method improves AUC‑ROC by 2.0% on BindingDB and 0.5% on DrugBank compared to prior biochemical representation learning approaches. Feature‑level analyses confirm the effectiveness of the probe-driven pattern matching, and the predictions are linked to an evidence‑guided downstream drug‑discovery workflow for candidate refinement and validation planning.
arXiv:2607. 18144v1 Announce Type: cross Abstract: Structure-based drug design (SBDD) leverages the 3D structure of protein targets, often complemented by other spatial constraints, to generate candidate binding molecules.
arXiv:2602.00663v3 Announce Type: replace Abstract: Optimizing molecules to achieve desired properties is a central bottleneck across the chemical sciences, particularly in the pharmaceutical industr...
MolDesignBench is a new benchmark for evaluating large language model (LLM)-based agents in scenario‑grounded molecular design. It contains 2,000 generation and optimization tasks that blend implicit narrative requirements with explicit property and functional‑group constraints, including infeasible cases, and require the use of 17 specialized chemistry tools. Experiments with leading LLMs show low success rates (best ~43%) and highlight failures in implicit‑constraint reasoning, infeasibility detection, and tool usage, underscoring the benchmark’s role in identifying key bottlenecks for future research.