SynLaD: Latent Diffusion for Generating Synthesizable Molecules Conditioned on 3D Pharmacophore Profiles

We present SynLaD, a latent diffusion framework for small-molecule generation that unifies ligand-based drug design objectives (what to make) with synthetic accessibility (how to make it). Current models typically optimize one objective at the expense of the other, creating a bottleneck for discovering high-scoring and synthesizable molecules. SynLaD combines reaction-constrained generation with pharmacophore-conditioned 3D design by learning a latent space that decodes to both 3D structures and synthesis pathways. An encoder maps molecules to a latent representation used by two decoder heads: (i) a geometric head that reconstructs atom types and coordinates and (ii) an autoregressive synthesis head that outputs synthetic routes in a serialized, reaction-based notation. A diffusion transformer generates novel latents in the learned space, conditioned on pharmacophore profiles. Across analogue generation tasks for bioactive ligands, SynLaD outperforms existing baselines in synthesizable and diverse hit generation, demonstrating that a single model can produce shape-aligned molecules with feasible synthesis plans.

Designing new small molecules for drug discovery requires balancing two goals: the molecules should contain the right chemical features to interact with a biological target, and they should also be possible to make in the lab. Many generative models can propose molecules with promising structural motifs, but these molecules are often difficult or impossible to synthesize, limiting their usefulness for experimental testing. SynLaD addresses this challenge by generating molecules through predicted chemical reaction pathways, so that each proposed molecule is associated with a plausible route for synthesis. At the same time, SynLaD conditions generation on three-dimensional pharmacophore features, which describe the spatial arrangement of key interactions needed for ligand-based molecule design. By combining 3D molecular design with synthesis-aware generation, SynLaD aims to produce candidate molecules that both match desired interaction patterns and are more readily testable. This can support faster design, validation, and iteration cycles in early-stage drug discovery.