SurfDesign: Effective Protein Design on Molecular Surfaces

A new protein design framework called SurfDesign models molecular surfaces as continuous geometric manifolds, integrating them with pretrained protein language models to improve functional protein engineering, according to a preprint posted to arXiv on 25 May 2026 [1][2]. The framework, detailed in a paper submitted to the quantitative biology section of the repository, addresses a limitation in current protein design: most methods condition predictions solely on backbone structure, ignoring the surface geometry and physicochemical complementarity that largely determine protein function [1][2]. SurfDesign employs surface-based equivariant message passing to capture features such as surface normals, curvature, and directional geometry, combined with a parameter-efficient fine-tuning strategy [2]. The authors report that SurfDesign consistently outperforms prior surface-conditioned and backbone-only methods on de novo binder and enzyme design benchmarks, and also achieves strong results on inverse-folding benchmarks used to assess structural compatibility [2]. The code has been made publicly available on GitHub [2]. The work appears on arXiv, an open-access repository that hosts electronic preprints across physics, mathematics, computer science, quantitative biology, and related fields [11]. As of November 2024, the repository was receiving approximately 24,000 submissions per month and had surpassed two million total articles by the end of 2021 [11]. The preprint is accompanied by experimental tools developed through arXivLabs, a framework launched in 2020 that enables community collaborators to build features directly on article record pages [10]. These tools include the Bibliographic Explorer for citation-tree navigation and the CORE Recommender for discovering related open-access papers [9][10]. arXivLabs collaborators operate under guidelines that require adherence to values of openness, community, excellence, and user data privacy, with access limited to minimal and anonymized user data [10]. Surface geometry has long been recognized as central to biomolecular interaction. In biomaterials science, surface modification at the molecular, cellular, and tissue levels is used to alter properties such as bacterial adhesion and cell adhesion while retaining bulk material characteristics [3]. The SurfDesign authors argue that manifold-aware surface representations offer a principled foundation for functional protein and enzyme design, extending this principle into computational protein engineering [2].