A new study published in Big Earth Data proposes an AI cube framework that integrates GeoAI models into geospatial data cube infrastructures to enhance large-scale Earth Observation data analytics. By introducing a model warehouse, intelligent model selection, and parallel inference pipelines on the Open Geospatial Engine platform, the approach significantly improves analytical capability and reduces inference time by over 80%. The framework advances the transition from traditional data cube processing toward AI-ready spatial data infrastructures.

At AACR 2026, Insilico will unveil four novel cancer inhibitors discovered via its end-to-end Pharma.AI platform. By harnessing trillions of data points and millions of molecular fragments, the platform integrates generative biology for target discovery with generative chemistry for de novo molecular design, accelerating the path from data to drug candidates.

Foundation models (FMs), which are deep learning models pretrained on large-scale data and applied to diverse downstream tasks, have transformed natural language processing and multimodal AI. However, in spatial transcriptomics (ST), no FM has yet demonstrated the capacity to generate novel, validated biological discoveries. The authors argue that this gap exists because ST data lack an explicit sequence-like structure, are noisy, and are more costly to collect than single-cell RNA sequencing data, making them unsuitable for simply reusing existing single-cell FMs. Therefore, how to leverage ST data to construct better foundation models is a highly promising research direction that warrants further exploration.