CellTune enables high precision analysis of spatial proteomics datasets through a human-in-the-loop active learning workflow.

Liu et al. develop Oncoformer, a multimodal transformer that reads routine laboratory tests and chest X-rays already collected in everyday care. Across more than 3.6 million individuals, it detects cancer, infers tumor stage, and stratifies treatment response and recurrence risk, pointing toward risk-adapted cancer care built on data already in hand.

Using plasma proteomics, Yu, Chen, Feng, Li et. al report a panel of five blood proteins that can predict future risk of metabolic dysfunction-associated steatotic liver disease (MASLD) up to 16 years...

UniPert-G2CP is an AI framework for jointly modeling molecular-level perturbation causes with phenotypic-level cellular effects across genetic and chemical domains. These complementary views enable more accurate perturbation prediction, mechanistic discovery, and efficient in silico drug screening.

Interactions between the tumor microenvironment and antibody drugs are imaged with spatial proteomics.

Mammals display prominent diversity in the ability to regenerate damaged ear pinna, but the genetic changes underlying the failure of regeneration remain elusive. We performed comparative single-cell ...

A prostaglandin prevents macrophages from clearing toxic neutrophils in aging organs

Studies of mice show that cancer dynamics depend on an interplay between the inherited genome and acquired mutations.

The cell-by-cell interactions between neoantigen-expressing tumor cells and tumor-specific T cells remained largely unexplored. A single-cell spatial transcriptomics framework enables mapping of neoan...

Spatially resolved transcriptomic technologies continue to enhance understanding of development and disease through a feedback loop involving academia and industry.

Multimodal machine learning reveals that tumour microenvironments can be decomposed into spatially organized multicellular ecosystems, termed spatial ecotypes, that can be accessed non-invasively via ...

The eORFeome, a large-scale collection of open reading frames encoding viral proteins and secreted bacterial and parasite effectors, enables functional genomics across diverse pathways and pathogens. ...

Hu et al. develop POND-seq, a nanocage-based strategy that exports RNA information from living cells without disruption. This approach enables repeated sampling of the same cell population to track transcriptome and RNA-protein interactions over time and supports scalable analysis of RNA-binding protein variants.

In this issue of Cancer Cell, Liang et al. report PathPrism, demonstrating how interpretable spatial representations of tissue organization can support biomarker discovery, clinical prediction, and hypothesis generation from routine histopathology slides. The framework highlights an emerging transition in computational pathology from black-box prediction toward biologically meaningful interpretation.

A radical new vision for healthcare: the full-body ultrasound. No radiation, no magnets, just sound, water, and 60 seconds.

An engineered ErbB oncogene-selective virus shows improved safety and efficacy for treating mouse models of ErbBr/HER2 positive ovarian cancer, showing promise for curing early cases and substantially...

Jin et al. developed an artificial intelligence tool that leverages histology to predict methylation-based central nervous system tumor subtypes and validated it for tumor classification, demonstratin...

In vivo CRISPR genetic perturbations are spatially mapped at scale.

Jia et al. develop SPACE-seq, a versatile technology that integrates CRISPR barcoding with spatial transcriptomics to jointly map cell ancestry, identity, and location. Compatible with diverse spatial platforms, SPACE-seq can be applied to a wide range of biological questions, revealing clonal architecture, migration, and fate-shaping interactions in development and cancer.

While genetic sequencing is routine in cancer care, translating a tumor's complex mutation profile into actionable treatment decisions remains a central challenge. MutationProjector is pre-trained fro...

Tertiary lymphoid structures (TLSs) are critical regulators of antitumor immunity, yet their spatial organization, maturation, and clinical relevance remain incompletely defined across cancers. We ana...

Zhang et al. present SPAC-seq, a sequencing-based spatial CRISPR screening technology capturing high-throughput perturbation libraries and spatial whole transcriptomes. Complementing this, TARDIS is developed as a statistical toolkit for spatial perturbation analysis. Together, they facilitate discoveries in spatial functional genomics, including cell localization, microenvironmental organization, pathway regulation, and ligand-receptor interactions.

Multimodal machine learning reveals that tumour microenvironments can be decomposed into spatially organized multicellular ecosystems, termed spatial ecotypes, that can be accessed non-invasively via ...

Sidewinder enables high-fidelity DNA assembly by separating the information that guides assembly from the final assembled sequence.

Multiplexed in-vivo perturbations, interpreted by a foundation model that has only ever seen human tumors.