graph states

Researchers have introduced a graph transformer model designed to improve RNA velocity predictions in single-cell genomics. RNA velocity estimates the future state of individual cells, but existing methods struggle with noisy data and complex cell trajectories. By applying transformer-based deep learning to gene regulatory graphs, the new approach aims to produce more accurate predictions of how cells develop and differentiate. If validated, it could strengthen research in developmental biology and disease studies, areas where precise modeling of cell dynamics matters.

Researchers publishing in Nature report a deterministic and programmable fusion technique for generating photonic graph states at scale. Photonic graph states are entangled light states central to photonic quantum computing, networking and repeaters. Making fusion deterministic rather than probabilistic, while keeping it programmable, is presented as a step toward scalable quantum photonics. The finding matters because scaling entangled photon sources has been a key bottleneck for optical quantum technologies.