Abstract:Background/aim: Multiple Myeloma is the second most common blood cancer, characterised by the accumulation of malignant plasma cells and the production of large amounts of a monoclonal immunoglobulin protein, in the bone marrow. The identification and progression/behaviour of molecular markers across stages remains a scientific challenge. This work aims to provide a holistic approach to the understanding of the disease progression, providing specific methodologies and candidate biomarkers, able to characterise and distinguish the disease state across stages. Materials and methods: Two large bulk RNA datasets were used to collect and integrate stage-specific information at the gene level by means of: (a) differential expression analysis to obtain differential expressed genes (DEGs), (b) a recently introduced computational methodology able to detect monotonically expressed genes (MEGs), (c) a proposed computational methodology that uses pairs of MEGs at sample level, to classify and discriminate different stages of Multiple Myeloma. Additional numerical metrics were applied to rank the performance of these pairs across samples, facilitating the characterization and differentiation of disease stages. Validation was conducted using five additional external datasets, which were then utilised to enrich the final selection of top-rated genes identified from the two bulk RNA datasets under study. The final top-ranked genes were further used for pathway enrichment analysis in order to provide candidate pathways per stage. Results: We first show that MEGs provide better statistics than DEGs, both at gene and pathway level analysis. Secondly, we show that the proposed computational methodology by means of MEGs reveals short lists of high discriminative genes across stages, which in turn highlight significant groups of pathways. Conclusion: We integrated traditional analysis of DEGs with a recently introduced methodology for identifying MEGs, creating a novel computational approach capable of identifying highly discriminative genes and pathways that can serve as candidate markers for stage identification in a single sample.

Identification of Monotonically Differentially Expressed Genes across Pathologic Stages for Cancers

Identification of Gene Expression in Different Stages of Breast Cancer with Machine Learning

Cancer prediction with gene expression profiling and differential evolution

Stage-specific gene ratios highlight genes and mechanisms related to presymptomatic and symptomatic Multiple Myeloma

Integrating Multiple Datasets to Discover Stage-Specific Cancer Related Genes and Stage-Specific Pathways.

Predicting Stage-Specific Cancer Related Genes and Their Dynamic Modules by Integrating Multiple Datasets

Identification of Cancer Development Related Pathways Based on Co-Expression Analyses

Identification of Differentially Expressed Genes Between Original Breast Cancer and Xenograft Using Machine Learning Algorithms

Gene-MOE: A sparsely gated prognosis and classification framework exploiting pan-cancer genomic information

Identification of Immune-Related Prognostic Biomarkers Based on the Tumor Microenvironment in 20 Malignant Tumor Types With Poor Prognosis

Tumor-Specific Gene Expression Patterns With Gene Expression Proffles

Pan-cancer Identification of the Relationship of Metabolism-Related Differentially Expressed Transcription Regulation with Non-Differentially Expressed Target Genes Via a Gated Recurrent Unit Network.

Colon cancer diagnosis and staging classification based on machine learning and bioinformatics analysis

MECoRank: Cancer Driver Genes Discovery Simultaneously Evaluating the Impact of SNVs and Differential Expression on Transcriptional Networks.

Global Differential Gene Expression in Cancers and its Implications for Building Robust Diagnostic Classifiers

Feature Selection for Breast Cancer Classification by Integrating Somatic Mutation and Gene Expression.

Identification of Gene Signatures and Expression Patterns During Epithelial-to-Mesenchymal Transition From Single-Cell Expression Atlas

Towards precise classification of cancers based on robust gene functional expression profiles

Identifying Disease Genes and Module Biomarkers by Differential Interactions

A Stochastic Model for Identifying Differential Gene Pair Co-Expression Patterns in Prostate Cancer Progression

Multi-Level Analysis and Identification of Tumor Mutational Burden Genes across Cancer Types