publications

2025

1. Thesis

   Protein Design at Library Scale

   Jacob Merle Gershon

   University of Washington, 2025

   Abs

   Doctoral dissertation on generative approaches to designing proteins at library scale.

2. bioRxiv

   Computational design of cysteine proteases

   Hojae Choi, Brian Coventry, Magnus Bauer, and 15 more authors

   bioRxiv, 2025

   Abs DOI

   De novo design of enzymes that use an activated cysteine nucleophile to hydrolyze the polypeptide backbone in a sequence-dependent manner, achieving rate enhancements over the background reaction of up to 3 x 10^7. The designs adopt folds very different from natural proteases, and crystal structures closely match the design models.

3. arXiv

   Hierarchical protein backbone generation with latent and structure diffusion

   Jason Yim, Marouane Jaakik, Ge Liu, and 5 more authors

   arXiv preprint arXiv:2504.09374, 2025

   Abs

   LSD is a two-stage generative model for protein backbones: latents are sampled and decoded into a contact map, then atomic coordinates are sampled conditioned on that contact map. The hierarchy enables new forms of control over generation while scaling to large datasets.

4. Nat. Biotech.

   Multistate and functional protein design using RoseTTAFold sequence space diffusion

   Sidney Lyayuga Lisanza, Jacob Merle Gershon, Samuel W. K. Tipps, and 16 more authors

   Nature Biotechnology, 2025

   Abs DOI HTML Code

   ProteinGenerator is a sequence space diffusion model built on RoseTTAFold that generates protein sequences and structures simultaneously by iterative denoising, guided by desired sequence and structural attributes. We design thermostable proteins with varying amino acid compositions and internal sequence repeats, and cage bioactive peptides such as melittin.

2024

1. ICLR-W

   Integrating Generative and Experimental Platforms for Biomolecular Design

   Cheng-Hao Liu, Jarrid Rector-Brooks, Soojung Yang, and 7 more authors

   In ICLR 2025 Workshop Proposals, 2024

   Abs

   Workshop bringing computationalists and experimentalists together around generative machine learning for de novo biomolecular design and its experimental validation.