Leveraging single-cell sequencing to classify and characterize tumor subgroups in bulk RNA-sequencing data.

PURPOSE: Accurate classification of cancer subgroups is essential for precision medicine, tailoring treatments to individual patients based on their cancer subtypes. In recent years, advances in high-throughput sequencing technologies have enabled the generation of large-scale transcriptomic data from cancer samples. These data have provided opportunities for developing computational methods that can improve cancer subtyping and enable better personalized treatment strategies. METHODS: Here in this study, we evaluated different feature selection schemes in the context of meningioma classification. To integrate interpretable features from the bulk (n = 77 samples) and single-cell profiling (∼ 10 K cells), we developed an algorithm named CLIPPR which combines the top-performing single-cell models, RNA-inferred copy number variation (CNV) signals, and the initial bulk model to create a meta-model. RESULTS: While the scheme relying solely on bulk transcriptomic data showed good classification accuracy, it exhibited confusion between malignant and benign molecular classes in approximately ∼ 8% of meningioma samples. In contrast, models trained on features learned from meningioma single-cell data accurately resolved the sub-groups confused by bulk-transcriptomic data but showed limited overall accuracy. CLIPPR showed superior overall accuracy and resolved benign-malignant confusion as validated on n = 789 bulk meningioma samples gathered from multiple institutions. Finally, we showed the generalizability of our algorithm using our in-house single-cell (∼ 200 K cells) and bulk TCGA glioma data (n = 711 samples). CONCLUSION: Overall, our algorithm CLIPPR synergizes the resolution of single-cell data with the depth of bulk sequencing and enables improved cancer sub-group diagnoses and insights into their biology.

Meningioma transcriptomic landscape demonstrates novel subtypes with regional associated biology and patient outcome.

Meningiomas, although mostly benign, can be recurrent and fatal. World Health Organization (WHO) grading of the tumor does not always identify high-risk meningioma, and better characterizations of their aggressive biology are needed. To approach this problem, we combined 13 bulk RNA sequencing (RNA-seq) datasets to create a dimension-reduced reference landscape of 1,298 meningiomas. The clinical and genomic metadata effectively correlated with landscape regions, which led to the identification of meningioma subtypes with specific biological signatures. The time to recurrence also correlated with the map location. Further, we developed an algorithm that maps new patients onto this landscape, where the nearest neighbors predict outcome. This study highlights the utility of combining bulk transcriptomic datasets to visualize the complexity of tumor populations. Further, we provide an interactive tool for understanding the disease and predicting patient outcomes. This resource is accessible via the online tool Oncoscape, where the scientific community can explore the meningioma landscape.