Multiplexed DNA imaging uncovers promoter−enhancer interaction dynamics and effects of epigenetic modifications.

Engineered DNA recombinases efficiently and specifically insert genetic cargos without the use of landing pads.

Photovoltaic devices attached to immune cells travel through the blood to inflamed brain regions.

A biophysical model uses knockdown or overexpression data to infer splicing factor activity.

A growth-coupled biosynthetic pathway enables xathommatin synthesis in the soil bacterium Pseudomonas putida.

T cell immunity declines with thymic atrophy, but thymic cell grafts rescue T cell function.

Big pharma shares its machine learning models with biotechs, but awaits definitive data on success of artificial intelligence-generated drugs.

Transformer-based large language models are gaining traction in biomedical research, particularly in single-cell omics. Despite their promise, the application of single-cell large language models (scLLMs) remains limited in practice. In this Comment, we examine the current landscape of scLLMs and the benchmark studies that assess their applications in various analytical tasks. We highlight existing technical gaps and practical barriers, and discuss future directions toward a more accessible and effective ecosystem to promote the applications of scLLMs in the biomedical community.

Each year, Nature Biotechnology highlights companies that received sizeable early-stage funding in the previous year. Constructive Bio is using synthetic E. coli and non-canonical amino acids to encode peptides and proteins for a range of applications.