Search bioRxiv⌕ Search

Biology subjects

Castanon, I.

Publications and source records attributed to Castanon, I..

2 recordsLinked to original sources

Targeting Ligand Shape While Preserving a Warhead through Guided Diffusion and Adaptive Staged Growth

Fragment-based growth models exploit pre-existing structures but can become constrained by early autoregressive decisions and generally lack knowledge of the final target shape. Here, we tested whether geometric guidance applied to a pretrained DiffSBDD model could steer generation toward the shape of a ligand B while preserving a warhead from ligand A. One-shot guidance increased target-volume coverage but predominantly produced disconnected structures. Reformulating the task as smaller scaffold expansions substantially improved the recovery of connected anchor components. Target-directed selection and adaptive growth increments enabled progressive movement toward B, while beam search improved endpoint recovery. Soft-scaffold inpainting provided graded control over the mobility of the inherited scaffold while preserving the fixed warhead. Together, these results show that inference-time shape guidance can be converted from disconnected volume coverage into controlled staged molecular growth, although performance remains dependent on the geometric compatibility of the ligand pair.

bioengineering↗

A Conditional Variational Autoencoder with QSAR-Guided Surrogate-Weighted Fine-Tuning and Cross-Entropy Optimization for Targeted Antimicrobial Peptide Generation

Machine Learning frameworks have emerged as a promising tool for antimicrobial peptide design; however, generative models remain limited by two persistent problems: the limited availability of experimentally validated peptides and the circular dependency of the models. In this work we present a conditional variational autoencoder pipeline that addresses both limitations through a modular architecture that combines both binary and quantitative experimental data and implements a multimodal approach to externally guide the generation. A transformer-based encoder successfully generated a discriminative 64-dimensional latent space (test AUROC 0.968, F1 0.919) separating antimicrobial from non-antimicrobial sequences. This latent representation conditions a species-specific LoRA fine-tuned ProtGPT2 decoder through a scalar gating function, which generates balanced antimicrobial peptides through two different modes; prior and perturb, depending on their generation starting points. We introduced a Surrogate Weighted Fine-Tuning (SWF) ensemble to eliminate the circular dependency and a Cross-Entropy Method to explore and exploit the latent space, leading to successful antimicrobial peptide generation. The best candidates exhibited competitive physicochemical characteristics, a mean helical fraction of 0.874 (mean pLDDT 83.7), and externally predicted efficacy evaluated by APEX.

bioinformatics↗