Routine FFPE sections support clinically compatible single nucleus transcriptomics across six human cancer types

Tumor cellular composition is increasingly recognized as a key driver of treatment response and resistance, but comprehensive profiling remains largely research-based. This study presents a clinically compatible workflow combining single-nucleus RNA sequencing with AI-based cell annotation to profile routine FFPE tumor samples. Validated across 116 specimens and six cancer types, the approach was reproducible, compatible with archival and low-input biopsies, and supports the use of cellular composition as a potential biomarker in precision oncology.