iiinova — where your digital self keeps evolving

Stephen Lu

@matrixmaster.me

25 followers 12 following 8 posts

🇨🇦 | ai for bio | cs phd @berkeley_ai | prev @bighatbio, @mcgillu

Posts Replies Media

Stephen Lu @matrixmaster.me · 08/06/2026

Huge thanks to my coauthors @aakarshv1.bsky.social, Kohei Sanno, Jiarui Lu, @matsen.bsky.social, Milind Jagota, and @yun-s-song.bsky.social! We will be presenting at ICML. Preprint: arxiv.org/abs/2602.18982 Code: github.com/thematrixmas... Blog: songlab-cal.github.io/cosine (8/8)

000

Stephen Lu @matrixmaster.me · 08/06/2026

The resulting samples remain structurally plausible and human-like. We think this is a promising step toward controllable evolutionary protein design: not just generating sequences de novo, but guiding the processes that produce them. (7/8)

100

Stephen Lu @matrixmaster.me · 08/06/2026

We also explore CoSiNE as a design model. With predictor guidance sampling, we steer simulated maturation trajectories toward desired properties at inference time, biasing evolution toward antibodies with higher predicted affinity for specific antigens. (6/8)

110

Stephen Lu @matrixmaster.me · 08/06/2026

Importantly, our selection score beats strong antibody and protein LM baselines on zero-shot antibody VEP across binding and expression datasets. This suggests that learning germline-to-mature evolution adds signal beyond antibody-likeness alone. (5/8)

110

Stephen Lu @matrixmaster.me · 08/06/2026

CoSiNE achieves this by comparing two likelihoods: How likely is a mutation under the learned maturation model? How likely is it under neutral SHM? The difference gives a selection score: enrichment beyond mutation bias. This improves VEP over transition likelihood alone. (4/8)

100

Stephen Lu @matrixmaster.me · 08/06/2026

A key challenge is that antibody evolution mixes two signals. Some mutations are frequent because they are likely under somatic hypermutation. Others occur because the antibodies carrying them are favored by selection. For VEP & design, we want to separate these effects. (3/8)

100

Stephen Lu @matrixmaster.me · 08/06/2026

Traditional antibody language modeling assumes sequences are i.i.d. — ignoring the time-dependent process of affinity maturation. To address this, CoSiNE explicitly models transitions: how likely is a mature antibody y to arise from a germline precursor x over time t? (2/8)

100

Stephen Lu @matrixmaster.me · 08/06/2026

Antibody LMs learn what looks antibody-like, but not how selection turns naive germline antibodies into strong binders. @aakarshv1.bsky.social and I are excited to share CoSiNE, a model that learns this germline-to-mature process for variant effect prediction and antibody design. (1/8)

142