A transcriptome-based classifier to determine molecular subtypes in medulloblastoma

Abstract

Medulloblastoma is a highly heterogeneous pediatric brain tumor with five molecular subtypes, Sonic Hedgehog TP53-mutant, Sonic Hedgehog TP53-wildtype, WNT, Group 3, and Group 4, defined by the World Health Organization. The current mechanism for classification into these molecular subtypes is through the use of immunostaining, methylation, and/or genetics. We surveyed the literature and identified a number of RNA-Seq and microarray datasets in order to develop, train, test, and validate a robust classifier to identify medulloblastoma molecular subtypes through the use of transcriptomic profiling data. We have developed a GPL-3 licensed R package and a Shiny Application to enable users to quickly and robustly classify medulloblastoma samples using transcriptomic data. The classifier utilizes a large composite microarray dataset (15 individual datasets), an individual microarray study, and an RNA-Seq dataset, using gene ratios instead of gene expression measures as features for the model. Discriminating features were identified using the limma R package and samples were classified using an unweighted mean of normalized scores. We utilized two training datasets and applied the classifier in 15 separate datasets. We observed a minimum accuracy of 85.71% in the smallest dataset and a maximum of 100% accuracy in four datasets with an overall median accuracy of 97.8% across the 15 datasets, with the majority of misclassification occurring between the heterogeneous Group 3 and Group 4 subtypes. We anticipate this medulloblastoma transcriptomic subtype classifier will be broadly applicable to the cancer research and clinical communities.

Aberrant DNA methylation distorts developmental trajectories in atypical teratoid/rhabdoid tumors

“…To associate DNA methylation with expression, we calculated DE genes (adjusted p < 0.01, minimum two-fold change) between tumor types in RNA-sequencing data and three public microarray datasets. Microarray expression datasets represented all the MB subgroups with similar subgroup proportions as in the DNA methylation microarray data ( Supplementary Tables 1 and 3 ) (66). We used Venn diagrams to summarize genes that were similarly regulated in both microarray and sequencing experiments ( Figure 3A ).…”

OpenPBTA: An Open Pediatric Brain Tumor Atlas

“…For example, we performed subtyping of medulloblastoma tumors, for which only 35% (43/122) had subtype information from pathology reports. Among the subtyped tumors, we accurately recapitulated subtypes using MM2S (91%; 39/43) or medulloPackage (95%; 41/43) 60, 61 . We then applied the consensus of these methods to subtype all medulloblastoma tumors lacking pathology-based subtypes.…”

“…Medulloblastoma (MB) subtypes SHH, WNT, Group 3, and Group 4 were predicted using the consensus of two RNA expression classifiers: 61 and 60 on the RSEM FPKM data ( analysis module).…”

Robust classification using average correlations as features (ACF)

“…This can conceptually not be reduced as it can for ACF. 2 Therefore, our method brings additional flexibility in terms of time complexity and is particularly advantageous over the KNN classifier for large training sets.…”

“…While powerful classifiers have been successfully implemented for commonly studied omic types, such as DNA-methylation [ 1 ] or the transcriptome [ 2 ], a widespread problem associated with many emerging omics technologies such as proteomics and single-cell RNA sequencing (scRNA-seq) is the strong prevalence of missing values, which hampers the direct applicability of most classification algorithms. The number of missing values is often additionally amplified by the integration of multiple individual datasets which is a common strategy to add statistical power to a study [ 3 ].…”