A UC San Diego team applies the CANDiT AI framework to map gene networks in colon cancer, pinpointing the stress-response protein PRKAB1. Activating PRKAB1 with PF-06409577 restores CDX2 expression in cancer stem cells, driving differentiation and selective self-destruction.

Key points

• CANDiT AI framework analyzes >4,600 tumor genomes to identify differentiation therapy targets like PRKAB1.
• PF-06409577, a clinical-grade PRKAB1 agonist, restores CDX2 in colon CSCs, triggering differentiation and apoptosis.
• Patient-derived organoids validate selective CSC targeting and predict up to 50% reduced recurrence risk via a 50-gene signature.

Why it matters: This work demonstrates a scalable AI-guided differentiation therapy that selectively reprograms and eradicates cancer stem cells, promising more precise and durable cancer treatments.

A team led by CRG Barcelona deploys EPI-Clone, a targeted single-cell DNA methylation profiling method, to reconstruct clonal trajectories and quantify how blood stem cell diversity erodes with age, uncovering myeloid-biased clone expansion and its role in inflammaging.

Key points

• EPI-Clone integrates targeted single-cell CpG methylation profiling on the Tapestri platform to capture clonal barcodes across 230,358 cells.
• Aged mice and human donors exhibit up to 70% dominance of a few HSC clones, with a shift toward myeloid-biased hematopoiesis linked to inflammaging.
• Distinct CpG subsets reflect both differentiation stage and stochastic epimutations, enabling simultaneous lineage mapping and clonal barcode generation.

Why it matters: This approach enables precise tracking of HSC clonal dynamics, offering insights into inflammaging mechanisms and potential early biomarkers of age-related hematologic risk.