PharosAI announced a multi‑institution consortium to build multimodal cancer datasets on 10x Genomics' Xenium spatial platform, backed by £18.9 million from UK government Research Ventures Catalyst. The program will aggregate decades of archived NHS tumor samples into linked genomic, transcriptomic and imaging datasets paired with AI models to accelerate biomarker discovery and drug development. Complementing large‑scale data initiatives, researchers published PlasMAAG in Nature Biotechnology—a computational method that improves plasmid reconstruction from metagenomic samples using assembly–alignment graphs and contrastive learning. PlasMAAG addresses a technical bottleneck in microbiome genomics and will enable more accurate tracking of mobile genetic elements that matter for antibiotic resistance and synthetic biology.