[2603.26378] Generative Modeling in Protein Design: Neural Representations, Conditional Generation, and Evaluation Standards

Computer Science > Machine Learning arXiv:2603.26378 (cs) [Submitted on 27 Mar 2026] Title:Generative Modeling in Protein Design: Neural Representations, Conditional Generation, and Evaluation Standards Authors:Senura Hansaja Wanasekara, Minh-Duong Nguyen, Xiaochen Liu, Nguyen H. Tran, Ken-Tye Yong View a PDF of the paper titled Generative Modeling in Protein Design: Neural Representations, Conditional Generation, and Evaluation Standards, by Senura Hansaja Wanasekara and Minh-Duong Nguyen and Xiaochen Liu and Nguyen H. Tran and Ken-Tye Yong View PDF HTML (experimental) Abstract:Generative modeling has become a central paradigm in protein research, extending machine learning beyond structure prediction toward sequence design, backbone generation, inverse folding, and biomolecular interaction modeling. However, the literature remains fragmented across representations, model classes, and task formulations, making it difficult to compare methods or identify appropriate evaluation standards. This survey provides a systematic synthesis of generative AI in protein research, organized around (i) foundational representations spanning sequence, geometric, and multimodal encodings; (ii) generative architectures including $\mathrm{SE}(3)$-equivariant diffusion, flow matching, and hybrid predictor-generator systems; and (iii) task settings from structure prediction and de novo design to protein-ligand and protein-protein interactions. Beyond cataloging methods, we compare assumptions,...