Tag
CellWorld introduces a latent-space predictive pretraining approach for spatial transcriptomics foundation models, predicting latent representations of masked cells instead of reconstructing gene measurements. Across held-out datasets, even small variants outperform existing baselines on all benchmarks, showing that scaling and broad biological diversity improve transferability.
COAST is a context-aware differential learning framework for predicting spatial gene expression from H&E histopathology images, using joint absolute and signed differential regression to capture spatial relationships.
SpaCellAgent is a self-evolving LLM-based multi-agent framework that automates end-to-end spatiotemporal trajectory analysis from natural language queries, achieving over 40% improvement in analytical efficiency while maintaining expert-level performance.
Arc Institute's PerturbSpace enables high-throughput single-cell profiling of transcriptome, location, CRISPR guides, clonal relationships, and surface proteins from many samples in one day, using standard single-cell sequencing.
PerturbSpace is a spatial transcriptomics method that presses a tissue section onto a chip of barcoded microwells, using antibodies to tag cells with location codes before single-cell sequencing, achieving >90% confident spatial assignment.
Machine Learning at Berkeley collaborated with LatchBio to benchmark their AI agent's performance on spatial transcriptomics workflows, evaluating its ability to automate complex bioinformatics tasks.