Related Experiment Video
Updated: Oct 6, 2026
Transposon Mediated Integration of Plasmid DNA into the Subventricular Zone of Neonatal Mice to Generate Novel Models of Glioblastoma
Published on: February 22, 2015
A machine learning classifier trained on cancer transcriptomes detects NF1 inactivation signal in glioblastoma
Gregory P Way1,2, Robert J Allaway3, Stephanie J Bouley3
1Genomics and Computational Biology Graduate Program, University of Pennsylvania, Philadelphia, PA, USA.
Background:
We have identified molecules that exhibit synthetic lethality in cells with loss of the neurofibromin 1 (NF1) tumor suppressor gene. However, recognizing tumors that have inactivation of the NF1 tumor suppressor function is challenging because the loss may occur via mechanisms that do not involve mutation of the genomic locus. Degradation of the NF1 protein, independent of NF1 mutation status, phenocopies inactivating mutations to drive tumors in human glioma cell lines. NF1 inactivation may alter the transcriptional landscape of a tumor and allow a machine learning classifier to detect which tumors will benefit from synthetic lethal molecules.
Results:
We developed a strategy to predict tumors with low NF1 activity and hence tumors that may respond to treatments that target cells lacking NF1. Using RNAseq data from The Cancer Genome Atlas (TCGA), we trained an ensemble of 500 logistic regression classifiers that integrates mutation status with whole transcriptomes to predict NF1 inactivation in glioblastoma (GBM). On TCGA data, the classifier detected NF1 mutated tumors (test set area under the receiver operating characteristic curve (AUROC) mean = 0.77, 95% quantile = 0.53 - 0.95) over 50 random initializations. On RNA-Seq data transformed into the space of gene expression microarrays, this method produced a classifier with similar performance (test set AUROC mean = 0.77, 95% quantile = 0.53 - 0.96). We applied our ensemble classifier trained on the transformed TCGA data to a microarray validation set of 12 samples with matched RNA and NF1 protein-level measurements. The classifier's NF1 score was associated with NF1 protein concentration in these samples.
Conclusions:
We demonstrate that TCGA can be used to train accurate predictors of NF1 inactivation in GBM. The ensemble classifier performed well for samples with very high or very low NF1 protein concentrations but had mixed performance in samples with intermediate NF1 concentrations. Nevertheless, high-performing and validated predictors have the potential to be paired with targeted therapies and personalized medicine.
Insights
We identified that neurofibromin 1 (NF1) protein loss, not just mutation, drives tumors. Machine learning can predict NF1 inactivation in glioblastoma, guiding synthetic lethal therapies.
Area of Science:
- Oncology
- Genetics
- Computational Biology
Background:
- Neurofibromin 1 (NF1) is a tumor suppressor gene. Inactivation of NF1 function, through mutation or protein degradation, drives tumor development, particularly in glioma.
- Identifying NF1 inactivation is challenging as it can occur without genomic mutation.
- NF1 loss alters tumor gene expression, creating a potential biomarker for targeted therapies.
Purpose of the Study:
- To develop a predictive model for identifying tumors with reduced neurofibromin 1 (NF1) activity.
- To determine if machine learning can detect NF1 inactivation in glioblastoma (GBM) using transcriptomic data.
- To enable patient stratification for therapies targeting synthetic lethality in NF1-deficient tumors.
Main Methods:
- Utilized RNA sequencing data from The Cancer Genome Atlas (TCGA) for glioblastoma (GBM).
- Trained an ensemble of 500 logistic regression classifiers integrating mutation status and whole transcriptome data.
- Validated the classifier on independent datasets, including samples with matched RNA and NF1 protein levels.
Main Results:
- The machine learning classifier accurately predicted NF1 inactivation in GBM, achieving an area under the receiver operating characteristic curve (AUROC) of 0.77.
- The classifier demonstrated robust performance on both raw RNA-Seq and transformed microarray gene expression data.
- The NF1 score generated by the classifier correlated with measured NF1 protein concentration in a validation set.
Conclusions:
- The Cancer Genome Atlas (TCGA) database can be effectively used to train predictors of NF1 inactivation in glioblastoma (GBM).
- The developed ensemble classifier shows potential for identifying tumors that will benefit from synthetic lethal agents.
- Validated predictors of NF1 inactivation hold promise for advancing personalized medicine and targeted cancer therapies.
More Related Videos
Related Concept Videos
02:39lncRNA - Long Non-coding RNAs
01:05Cancer-Critical Genes II: Tumor Suppressor Genes
When the function of certain critical genes, especially those involved in cell cycle regulation and cell growth signaling cascades, gets disrupted, it upsets the cell cycle progression. Such cells with unchecked cell cycles start proliferating uncontrollably and eventually develop into tumors.
Such genes that act...
01:05Cancer-Critical Genes II: Tumor Suppressor Genes
When the function of certain critical genes, especially those involved in cell cycle regulation and cell growth signaling cascades, gets disrupted, it upsets the cell cycle progression. Such cells with unchecked cell cycles start proliferating uncontrollably and eventually develop into tumors.
Such genes that act...