Diagnostic classification of childhood cancer using multiscale transcriptomics

Abstract

The causes of pediatric cancers’ distinctiveness compared to adult-onset tumors of the same type are not completely clear and not fully explained by their genomes. In this study, we used an optimized multilevel RNA clustering approach to derive molecular definitions for most childhood cancers. Applying this method to 13,313 transcriptomes, we constructed a pediatric cancer atlas to explore age-associated changes. Tumor entities were sometimes unexpectedly grouped due to common lineages, drivers or stemness profiles. Some established entities were divided into subgroups that predicted outcome better than current diagnostic approaches. These definitions account for inter-tumoral and intra-tumoral heterogeneity and have the potential of enabling reproducible, quantifiable diagnostics. As a whole, childhood tumors had more transcriptional diversity than adult tumors, maintaining greater expression flexibility. To apply these insights, we designed an ensemble convolutional neural network classifier. We show that this tool was able to match or clarify the diagnosis for 85% of childhood tumors in a prospective cohort. If further validated, this framework could be extended to derive molecular definitions for all cancer types.

Genomic Newborn Screening for Pediatric Cancer Predisposition Syndromes: A Holistic Approach

“…Compared to traditional methods, newborn genomic screening has the potential to identify genetic predispositions to certain pediatric cancers at an earlier stage, allowing for earlier intervention and potentially better outcomes [ 181 , 182 , 183 ]. It is important to note that the cost of newborn genomic screening may decrease over time as technology advances and becomes more widely adopted [ 184 ].…”

A Review of Single-Cell RNA-Seq Annotation, Integration, and Cell–Cell Communication

“…Additionally, current clustering algorithms often fail to incorporate the intrinsic hierarchical structure among cells, leading to potential inaccuracies. However, the recent introduction of RNA-seq clustering techniques that integrate biological realities into their models [38] holds promise for the future development of multi-level, multi-scale clustering strategies that are tailored specifically to scRNA-seq analyses.…”

“…Interestingly, a similar approach based on multi-level scale-adaptive clustering has been reported for the unsupervised classification of tumor subtypes using RNA-seq. This approach, known as Resolution-Adaptive Coarse-to-Fine Clusters Optimization (RAC-COON, [38]), classified more than 13,000 samples into an eight-level hierarchical tree based on their expression similarities. It successfully generated an atlas consisting of 455 tumor and normal classes.…”

Overview of modern genomic tools for diagnosis and precision therapy of childhood solid cancers

“…Recently, Shlein et al [ 8 ▪▪ ] reported an intriguing method based on RNAseq data to molecularly define most childhood cancers and accurately predict subgroups and corresponding outcomes. Their methods measured transcriptional entropy and demonstrated significant diversity both between and within tumor types, in contrast to the relatively quiet genomic DNA landscape of most pediatric cancers.…”