Multi-omics biomarker selection and outlier detection across WHO glioma classifications via robust sparse multinomial regression
Carrilho, J. F.; Coletti, R.; Costa, B. M.; Lopes, M. B.
Show abstract
Gliomas are aggressive brain tumors difficult to treat mostly due to their large molecular heterogeneity. This requires continuous improvement in the molecular characterization of the glioma types to identify potential therapeutic targets. Advances in glioma research are rapidly evolving, contributing to the updates of the WHO classification of tumors. Data analysis of multiple omics layers through classification and feature selection methods holds promise in identifying crucial molecular features for distinguishing between glioma types. We developed a robust and sparse classification workflow based on multinomial logistic regression to investigate the molecular landscape of gliomas. We considered transcriptomics and methylomics glioma profiles of patients labeled following the latest WHO glioma classification updates (2016 and 2021). Overall, our results show a notable improvement in glioma types separability for the 2021 WHO updated patient labels at both omics levels. Patients flagged as outliers for the 2016 WHO classification exhibited a molecular profile deviating from the one of the respective classes, which was more aligned with the current associated glioma type according to the 2021 WHO update. The methylomics profiles were particularly promising in the identification of outliers. These contributions will support further revisions of glioma molecular characterization and the development of novel targeted therapies.
Matching journals
The top 6 journals account for 50% of the predicted probability mass.
Similar papers in this journal
- A Novel Tool for Multi-Omics Network Integration and Visualization: A Study of Glioma Heterogeneity 98%
- Development of an absolute assignment predictor for triple-negative breast cancer subtyping using machine learning approaches 95%
- A Comprehensive Targeted Panel of 295 Genes: Unveiling Key Disease Initiating and Transformative Biomarkers in MultipleMyeloma 95%
Similar papers in this journal
Similar papers in this journal
- Imputing missing RNA-seq data from DNA methylation by using transfer learning based-deep neural network 95%
- Disease classification for whole blood DNA methylation: meta-analysis, missing values imputation, and XAI 94%
- MLcps: Machine Learning Cumulative Performance Score for classification problems 93%
Similar papers in this journal
- Mime: A flexible machine-learning framework to construct and visualize models for clinical characteristics prediction and feature selection 94%
- Benchmarking feature selection and feature extraction methods to improve the performances of machine-learning algorithms for patient classification using metabolomics biomedical data. 94%
- Topological embedding and directional feature importance in ensemble classifiers for multi-class classification 93%
Similar papers in this journal
- DeepInsight-3D for precision oncology: an improved anti-cancer drug response prediction from high-dimensional multi-omics data with convolutional neural networks 95%
- Tensor decomposition- and principal component analysis-based unsupervised feature extraction to select more reasonable differentially expressed genes: Optimization of standard deviation versus state-of-art methods 94%
- High tissue-specificity of lncRNAs maximises the prediction of tissue of origin of circulating DNA 94%
"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.